Returned

I think that the issue is mostly with movement. Shardmetal is very rigid and as far as we've seen is very difficult to manifest as anything other than one piece. So even though you might be able to manifest the spren as something armor-shaped it would probably be just one single, inflexible piece, which would make moving very difficult. It could work well as a shield, like Kaladin used in WoR, or armor that works better as a single piece (like a breastplate).

Even where there is a clash of intents, it may not be as direct or immediate as between Preservation and Ruin. I think that the issues are twofold: Combined Shards become one, single thing which is different from the two separate things they were before. This was what Rayse wanted to avoid: being something other than Odium alone. Whether or not there is conflict is a different question, but it seems reasonable to think that at some point there is something that Retribution might desire or pursue that is different from what Odium would desire or pursue. I don't think there is any indication of how severe such a thing needs to be, though. The Vessel needs to direct the Shard (or be consumed by it), and there are many ways that this direction could be pursued. For example, Sazed could be Harmony or Discord, and failing to maintain a good balance as the former may lead to the latter-- not one Shard winning out over the other but a combination that may not be what the Vessel would prefer or could use effectively in pursuit of their goals. We've seen some consequences of Sazed's struggles after a couple of centuries of the tension between his Shards but have seen only a very brief period of Taravangian's. Taravangian's opinions of his nature with his two Shards may not be accurate nor sustainable, and he may get better or worse at handling the situation as the Shards influence his mind and will. I suspect that issues satisfying Honor and Odium together will eventually be meaningful for Taravangian, but we'll see how well he manages. The Shards' natures are not quite innate. They were all one, single combined thing at one point, and the sixteen specific pieces are sort of arbitrary divisions of that one thing as effected by the Shatterers. So far we know very little about the specifics but "two Shards" might be a less accurate way of thinking about it than "a single Shard which needs to find its own equilibrium, which may not be deterministic". Given the effects we've seen from holding just one Shard I am skeptical that adding another can be fundamentally imbalanced in the way that "Odium suppresses Honor" would suggest, though maybe a stable balance between the two would look a lot like that to us.

Ba-Ado-Mishram is a spren and the Heralds are not. Other differences may exist, too, but that one seems relevant (given what we know so far). I don't think that the gem killed Jezrien so much as it failed to contain his cognitive shadow which, without the Oathpact preserving its existence, faded to the beyond. The Raysium knife is what killed him and the gem was irrelevant to the process. BAM is her own self-sustaining thing and is not generally at risk of slipping into the beyond.

As I recall, aluminum blocks Investiture but doesn't sever connections: an allomancer inside of an aluminum cell can still burn metals, they just can't project the effect of that burning through the aluminum. The aluminum surfaces don't block the connection to Preservation which makes Allomancy possible. This should be like what happened to Chana, as the aluminum doesn't sever the connection to Honor/the Oathpact. The Raysium knife is different, actively draining the Investiture into the gemstone (as cognitive shadows, this Investiture is most of what the Heralds are) and apparently severing the connections to the Oathpact which sustained them. Had the cognitive shadow stayed contained in the gemstones they might have been imprisoned indefinitely, like a spren, but that didn't happen for them. Instead the connection which allowed the Heralds to persist after death was gone (or cognitive shadows don't behave like spren in this way), so death was permanent and the gemstone did not contain them. The key difference seems to be the action of the Raysium.

That's fair, and I think that Cosmere books generally have largely exhausted the goodwill for it. It does become problematic when we demand an explanation for why every possibility was rejected in favor of what was actually chosen. Windrunners should absolutely be dropping Shardnets with weighted edges from on high, re-summoning them, and repeating indefinitely. It stings more for Cosmere books because so much effort has gone into the low-level details of how magic systems work, and larger-scale applications of that same attention would produce interesting situations. Conventional medieval-style combat should be rare and seriously disrupted after OB at the very latest. But there are endless reasons why one tactic might be preferred over another, or disfavored altogether. The books are bloated enough without expanded 10-fold to detail things that specifically don't happen.

Sure, but we also get only a tiny fraction of the details relevant to in-world situations. If you'd rather assume we have complete and reliable information, meaning that these gaps are only arbitrary and sneer at the books because of it, have at it. If someone would prefer to acknowledge that our information is incomplete and that some of what we don't know could plausibly explain gaps we perceive, I think that's fine too. That we've never seen a soulcasting on the scale of the dome, and that it might require a high-Oath Radiant and a direct conduit to the SR (which Azimir didn't have) to accomplish, is enough for me to not automatically declare that situation to be sloppiness alone. I agree that the Windrunners should be a lot more devastating than they seem to be and it stretches credulity that they aren't. Maybe part of that is some reality they face I'm not thinking of and that the books don't mention. Maybe some of it is detail that's expressed in the books in a wobbly way, like how far an amount of Stormlight goes. Maybe the scales of battles do involve that many casualties, hard as that is to believe. Maybe some of it is lack of actual experience with tactics on our parts. Maybe it's just arbitrary plotting and lack of thinking through the details. But even if the lattermost is true of the Windrunner's performance, that isn't evidence that some unrelated thing is also equally arbitrary and careless.

This is an excellent example, thanks for bringing it up! I put it up there with Kelsier's death in Final Empire. I think that the reason these feel better to me than most others, like Teft or Wayne, is that it's very specifically something that is about their stories and desires. Kelsier and Lightsong knew that they were likely to die and made the choices that led to that end anyways-- they fulfilled their stories. Teft's death was wholly an object of others' stories. Wayne's death didn't hit me too hard because nearly all of his characterization came in the final volume of his stories, nearly all via flashback, and it really felt like the whole situation was one Harmony placed him into. Teft and Wayne barely had their own stories, though I concede that they had backstories. There's less drama in being a sidekick or an extra than there is in having a big, detailed story that is largely about you. I hadn't thought of it this way before, and my initial reaction is that I'm not sure I agree. The Jesus allegory sort of covers any self-sacrificing death, which in turn reflects heroic commitment on the hero's part and the gravity of the threat they're resisting. Characters like that tend to be the most major of characters and it's an incredibly common trope. Other characters' deaths tend to matter to readers based on their affinity for those characters. A sudden, incidental death or helpless execution is always kind of a pathetic moment, devoid of drama in itself. I liked Teft well enough but his storyline of overcoming his past failures and shame is literally the same as every Radiant's story, and he mattered to the story mostly because of his relationship with Kaladin. If he'd been a more realistically-portrayed addict I don't think I would have cared more about his murder just because of that. If he'd been a more rounded character, with more of his own goals, ambitions, experiences, and relationships shown on screen, that probably would have done it. His POV chapters gave us more than nothing (I cared more about his death than the countless, nameless soldiers we've seen fall). But it was a thin sketch that, for me, didn't produce much of a connection. He didn't have anywhere near the development or richness of a Kaladin or Kelsier. I feel like there's a double-edged sword with representation. I thought Renarin was interesting and well developed enough for a minor character in the earlier books, and his representation there was sufficient for a lot of readers to identify his autism long before it was confirmed. Readers could see things in him and identify with those things. But after he got the designation officially we got lists of things that are unmistakably diagnostic criteria, and jamming those in felt less natural than his previous expressions of relevant traits. His sexuality was similar, and he spent more words being explicitly described with unmissable statements than he did being an actual character who happens to be queer in ways that he may demonstrate while living his life. I don't think that trend is specific to Sanderson, the zeitgeist seems to really value explicit taxonomy and some externalization of self. I'm not intending to criticize that, exactly, but I do think that it tends to make for worse prose stories.

It's a conclusion, and I won't bother trying to dissuade you (I'm not impressed by their competence either, though my assessment is less harsh than yours). It will cover all possible issues and all argumentation against it is pointless, since they can just be too stupid and incompetent ever to accomplish anything. And if you start with that insistence and then use it to backfill every possible explanation then thinking about details is also pointless; the incompetence is really the only thing that matters. Your insistence that there is an "easy win" button that literally everyone in the books is too stupid to press, to the point that you refuse to even consider that such a button might not exist, is odd given that we know our information is so incomplete. But you seem convinced and committed. That's not the same, however, as assuming that some specific tactic definitely would have worked and the only conceivable point of failure is the characters' boundless incompetence. If nothing else the cost in Light and gemstones could easily be enough to render the dome unworkable even if they had a Soulcaster otherwise capable of creating it (I guess the incompetence argument doesn't include their ability to Surgebind). While we don't know much about the full range of powers the Fused have, Ash and Taln certainly do (or at least know vastly more), and those details might be enough to suggest that a given strategy is impractical or won't be effective. Odium thought he would win with the forces he allocated given the situation, and was wrong. Someone is going to win, and it turned out to be the defenders, though it was close. Any change in circumstances on the ground might have changed things in ways that widened that margin of victory but might also have narrowed it, even to the point of a loss. The books must have become very unsatisfying for you. As someone who is less satisfied with the recent releases than the earlier ones, I sympathize.

Sanderson does a really good job of setting up interesting situations but then lets them fade as he moves on to the next thing. I think that he likes the worldbuilding and enjoys the options those situations offer but doesn't consider them to be where the story is and so doesn't feel a lot of pressure to keep them up. This is reflected in situations not being resolved or otherwise paying off (such as the tensions in the Elendel Basin) and in being tidied up so that he can get back to the "real" plot (like the Urithiru takeover or the Rosharan coalition). In those respects I think it's a matter of scale: the major plots of the books have interstellar and interdimensional scope, so it's easy to ignore something like a regional tension and hard to make them both matter equally at the same time. A lot of them are also hard to resolve in a way that is both quick and satisfying: deflating regional tensions on the brink of civil war seems hard to do in a handful of pages, especially when it's mostly in the background. When these sorts of details are used as flavor I get kind of frustrated when they end up not just ignored, but overridden. Alethi culture, especially as distinct from others on Roshar, is a big deal and comes up often in WoK and WoR: the Alethi way of doing things governs and describes a lot of behaviors and provides constraints on what certain characters can do in certain sitations. But it fades quickly afterwards, and Alethi end up behaving in pretty much the same way as Thaylens or Kharbranthians. It seems like it's too much bother for Sanderson to keep dealing with, so he just ignores it. It makes a lot of the writing feel thinner to me, like there are fewer details that matter and the setting is less than it used to be. I'm a little bit more tolerant of characters. They take a lot of time and plot capacity to develop and you can't guarantee that people will respond to them in the ways you would like. You can't reliably just spin up a new Teft or Kaladin on demand, so killing one off (or changing them significantly) is a really big deal. With how long and expansive the Cosmere story is we should be seeing more major characters fall but I think that, along with the difficulty and risks involved in bringing in new characters, the considerations are: fan service (fans will not tolerate their favorites being killed very well and are unlikely to be satisfied by a death being thematically appropriate or necessary), and it seeming excessive and arbitrary (a common Game of Thrones critique). It's a hard needle to thread with so many books written and yet to be written, especially with so many magical options that make death a more surmountable problem. Even so, I agree that the fake-out deaths are a real problem because they are still deployed to get the full emotional impact a real death would elicit while also feeling arbitrary and cheap when reversed. It's gotten to the point that I barely respond to a major character's death because I think it's unlikely to stick. Their non-death consequences tend not to stick as much because many of the stories are about their personal growth (so they overcome or move beyond those consequences one way or another), and they are often chances to show off specific details of applied magic systems. I think that Sanderson can commit to negative consequences but agree that he generally doesn't. He reserves too many outcomes for the future, saves himself the trouble of developing negative outcomes and how characters deal with them when he doesn't care about scenarios any more, and keeps characters focused on conflicts they can't change very easily to maintain constant tensions over a lot of text.

I won't go into a ton of detail to avoid spoilers for the last book, but I agree that much of the tower was rushed and shortchanged. Many of the events of the final books felt rushed to me. I always felt that Noal was neglected as a character, and his end continued that. Some of what you describe is thematic, I think. A mundane obstacle is less important than the extradimensional rules that humans can't understand nor account for, and detailing things that don't work like a human would expect they should seems like it could get old pretty quickly. There is only so much strangeness that can be described before that alien quality starts getting lost. But a maze that fundamentally can't be solved does feel like an underwhelming gimmick.

Aren't Regals also superhuman, just less so than Fused? And if conditions in Azimir required a different mix of troops it's plausible that the mix of troops across the fronts would have changed. The idea that the defenders could have radically altered the situation (if they even could in this way) but the invaders would not have altered their approach at all seems unlikely. Maybe some of the Unmade would have been deployed instead of ignored. What they had to do, and the tools they needed to prepare, depend on what they're going to do. A dome (or more realistically, wall) changes the needs of the invasion, to which the invaders will respond, as above. Sending Nale with some of his Skybreakers would have provided Shardblades, as a trivial example. Shovels, Archimedes screws, armored claws, Thunderclast emergences? Tunneling is hardly a new innovation and is not unapproachably difficult. And even though a breach offers some advantages to the defenders, the invaders were already committed to significant losses in assaulting the city-- it's not all that different from the defenders surrounding the Oathgate, but their forces would be more stretched out around the dome's greater circumference. I'm not saying it wouldn't suck for the invaders, I'm only saying that it's not the automatic, easy win which would automatically justify the effort (assuming, again, that it's even possible). We haven't seen all of the Altered Ones' powers (at least, I don't think so). Their external use might be similar to an Elsecaller or Lightweaver's souclasting ability, and would be worth the cost in Voidlight if it were the only way to overcome a static fortification. They could also just spit the appropriate acids until they've dissolved/made brittle/otherwise degraded or overcome the bronze at as many points are are necessary. If they did it high up on the dome/wall they would have high ground surrounding a fortified staging area from which to launch their sorties and assaults. The Elsegates "tap into spiritual realm shenanigans" and permit faster-than-light travel, which seems irrelevant to a realm crossing and therefore implies some ability to cross physical distances. But it doesn't really matter, as I bring it up mainly to demonstrate that the defenders don't know everything about what their enemies can do which makes total confidence in a static defense a less than awesome idea. It doesn't even have to be that. It could be as simple as redirecting Heavenly Ones to fly soldiers to a different place. Even though it's expensive in Light and logistical ability it might be worth it if the dome is as insurmountable as you insist. Or sending in Thunderclasts. Or sending in the Unmade. Or using their skills and abilities to deal with the dome, possibly even trivializing it. The defenders don't know everything but need to win, have limited resources and time, and so an extreme dedication of their resources is a big commitment. The dome is already dubious in terms of whether or not they could have done it at all, which I think is the major impediment, but even if we were to grant that it was possible it's not the perfect defense you have presented while also being very expensive. Maybe still worth trying, but also maybe not. Prohibitively expensive, not effective enough, impossible to accomplish, or any combination of those: it's enough to grant that the characters rationally chose to pursue something else, unless you really want to conclude that they're all stupid and incompetent and the sequence is just badly conceived. Take your pick at your preference, but it's not an inescapable conclusion.

Retribution's main focus (once the time dilation is over) is likely to be on the broader Cosmere rather than Roshar, or at least it seems so to me. He has to worry about proactive interference from other Shards, his plans all involve expanding into other places, and I imagine he feels like he's won pretty handily on Roshar. I think that the initial stages of the back five books will be the Radiants dealing with the control Retribution's followers will have established, increasing their ability to act and ending their siege-like situations. Only later on will they be contending directly with Retribution again. Opposing Retribution will be difficult. I think that his greatest weaknesses are that he is overconfident in his abilities (especially now that his emotive reasoning is backed by Odium's influence, and his intelligence has some flavor of finding arguments to justify his emotional conclusions rather than independent, rational reasoning), his fondness for gloating over his opponents, and the independence of Honor. Odium degrading his will and mind will cut against his scheming, and Honor having a mind of its own seems like it will be hard for him to deal with as he pursues what he wants rather than what his Shards demand. I suspect that his ultimate defeat will involve not being able to dominate his Shards as he has been able to in the early days of his ascension, though whether or not he is made more impotent (like Sazed), is affected by adding another Shard to his mix, or he is replaced by another Vessel as the Cosmere's game of musical Shards continues I cannot say. It seems likely that this will be tied in with the Cosmere's ultimate conclusions, though at this early point who knows? After WaT my thinking has been that ultimately all of the Shards are going to be separated from Vessels until they start to gain their own sentience, like Honor has, removing the human pettiness and failure from the picture and leaving the underlying forces of the Cosmere untampered with. Whether or not they're recombined again I don't know, though I lean towards yes.

I'm not so sure. Odium's forces eventually built their own fortifications within the Oathgate platform (I think? I really, really need to reread WaT). A wall can aid the attackers in a variety of ways, for example giving them an opportunity to stage their forces in the physical realm and sortie at their preference, or hiding what they are doing from the defenders. Maybe it would have helped the defenders and maybe not, but I don't see a dome or wall as necessarily being all that helpful for them. Especially if it's not so easy to create. Even a knee-high wall seems dubiously useful, given the physical abilities of the Fused. Another instance in which we don't know enough details about how the power works, but as noted above by @Nameless and @alder24 it's clearly not just stone, whatever the limitations turn out to actually be. Sure, but some of their options are faster than others. A light-antilight explosion might work quickly, a Shardblade will let them carve tunnels more quickly than any mundane method. They could even just tunnel under it. They also have access to Soulcasting in at least some forms, which offers options to dissolve or deform the bronze in ways that make it less of an obstruction even if they can't just cast it into nothing (or, worse, something dangerous to the defenders). They only need one accessible, effective way to deal with the wall and then its more of a hindrance to the defenders than it is helpful. El did it at the Shattered Plains, uninteresting as it was. That the characters don't know is exactly the point: they know they have incomplete information, and so in at least some respects knowing that the invaders will come through the Oathgate is valuable because they can plan their defenses around that approach. If forces fly in, Elsegate in, or whatever else, the defenders will be more surprised and less prepared, and even if they could make it their wrong-way-facing Maginot Line dome will be worse than useless because they spent their very limited time and resources to create it rather than doing something else.

I'm not sure of the dimensions of a barrack relative to the Oathgate, nor the thickness of the walls. I feel comfortable assuming that a dome covering the Oathgate would be massively taller than any barracks on the Shattered Plains, and the walls certainly aren't 15 feet thick. The cost in Light is also a factor, though also a frustratingly unclear one in this application. Gemstones crack, available Light is consumed by Radiants, it's not unreasonable to think it might be possible that there would not be enough resources to perform such a large-scale task. Not knowing the scale makes it hard for me to be fully confident in saying this definitely is so, but it seems reasonable to me that this might be more than they could confidently do in the time available. Do you have a citation on this? My understanding is that they can move through any solid object that was never living matter, not that they have some special ability to interact with stone only. If you have information more precise than that I'd appreciate the chance to learn it. The Deepest Ones were just an example of the dome not being impervious, there are other ways it could be melted through, chipped through, blasted apart, or carved through with Shardblades. Or Odium's forces might have just come through another way, if the Oathgate were made prohibitively expensive to deal with.

I thought that the Tower of Ghenji sequence was OK, at least in that Mat's luck was what made the difference for them-- no mortal preparation nor planning that would work in the the "regular" world is enough to navigate it, which is how Birgitte's attempt failed so badly. I agree that events there felt too condensed, and that Noal's death (and decisive revelation as Jain) felt rushed. I also felt like most of the strangeness of that realm had already been covered and there wasn't a whole lot new about it despite the characters seeing and experiencing more of it. I guess it's hard to keep the bizarre, otherworldly feel while also demonstrating a lot of it. Definitely it was not worthy of the buildup, since Mat's luck was the only really important element and that had already been very well established and solved a lot of problems.

These two statements contradict: if it can be done, it can be done, and since we know it was done, it can be done. Unless I've missed or am forgetting something definitive and official I think we should be cautious in assuming we know enough about the mechanics of how the lerasium and atium were made that we can broadly define things as impossible. There are other differences between Harmony and Retribution, perhaps most relevantly that Honor is more independent than Ruin or Preservation (maybe enough to matter for something like this). Either way the process might be impractical-- I doubt Roshar will be importing others' godmetals just for this. Maybe Ba-Ado-Mishram can supply Voidlight somehow, a she apparently did in the past, though with the Well of Control having changed maybe she no longer can. Regardless, it does seem likely that the Fused can use Warlight to fuel their abilities so a lack of Voidlight might not be important. We also might see something different, like Voidbinding rather than Surgebinding (if Voidbinding differs in how it's fueled, which it may not; we know so little about it). We could even see something more dramatic and exotic, like a new magic system of Warbinding or something. Sazed changed Allomancy on his ascension, so it seems likely that Taravangian can also change his constituent magic systems (maybe in this way, maybe not).

I think that Retribution probably could still produce Voidlight or allow it to exist, similar to how lerasium and atium can still exist and be produced after Harmony's ascension. But it's not really in Retribution's interest to continue with Stormlight since it's a major resource for his enemies, while Voidlight (or Warlight) might be just a good for his own forces. This quote itself indicates that Odium knows where the perpendicularity is: he doesn't trust anyone to know its location, suggesting he could tell but chooses not to, not that he is incapable of telling anyone because he, himself, doesn't know. Whatever the reason it was placed on the Shattered Plains, it might still be important enough that it be there even if the Listeners maintain or control terrestrial access to it. Maybe he'll move it anyways (it doesn't seem like all that big of a deal to form one after Lost Metal, at least). The Listeners aren't very lorewise compared with the Fused, so maybe if it's not hidden it's better that they control access to it than that Fused like El be able to.

I think that the major issues are that it may not have been possible (that's a lot of soulcasting to do, especially all at once), it may not have been effective (the Fused likely have options for dealing with the bronze, including but not limited to Deepest Ones, and might not have been delayed much at all), it creates a fortified beachhead for the Fused invaders to use against the defenders (they eventually built their own fortification in the Oathgate complex, though on a smaller scale), and it changes the strategic picture (they know the point of invasion and general invader strategy with the Oathgate open, but if sealed this way the Fused might have done something else; they would not have left Azimir alone). The in-universe explanation for why they didn't do it is because the characters know it wouldn't have worked, for one reason or another (even if that reason isn't expressed on the page). The narrative reason is that it was important that Azimir be a major battleground, so the invasion would have proceeded somehow no matter what the humans did. Fortifying around the Oathgate to resist the invaders is what the defenders did anyways, so it's not quite a total strategic oversight, and they were successful in the end (well, kind of...). I, personally, am satisfied with an explanation that the soulcasting (even ignoring the cost in Light and adequate gemstones to actually do it) would have been prohibitively difficult to do with the time and resources they had available.

I think there's some magic involved in Koloss metabolism and sustenance. It's hard to think of a human-scale animal body being sustained by grass and ash, never mind a super high-performance body like a Koloss. If they can cheat the other realistic constraints for nourishment such that ash and dirt are enough, it doesn't seem plausible that they couldn't cheat enough for sand to also be enough. Not that I'd be surprised to see that limitation, I just don't think that "it's not realistic because of the chemistry" is a standard we can rely on for this.

She and Odium both seemed to expect that he would attack her, and I think Odium specifically said he'd destroy her when Dalinar was asking about the consequences of releasing him. I don't think that in general Shards' attacks on each other need to be invited/permitted by some pre-existing direct action, there were just extra restrictions on Roshar with these three specific ones. If Odium is released from those, then he goes on destroying other Shards.

There isn't any real evidence to suggest that she isn't a Herald, but I don't see any particular reason to think that she is one, either. Roshar has lots of regular assassins running around (or at least Rosharans of note act like it does), and blending in is a pretty common skill for such a person to have (such as Sadeas' carpenter, whom no one has accused of being an offworlder). If anything I would think that the Weeper's method of killing and mutilating the corpses suggests that she isn't a Herald or Worldhopper. A Herald has so many abilities that would make them irresistible assassins without a Shardblade that I don't see any compelling reason for them to use one, and when she takes such care to hide that a Shardblade was used at all what would be the point of all that bother? Their superhuman speed alone is more than enough even without any weapon. Other Worldhoppers tend to have other magic systems available which are even less obvious to a Rosharan than a Blade would be. Conversely, having an unknown or lost Blade is an amazing tool for an otherwise ordinary Rosharan assassin: it can't be perceived unless she is actively using it, can't be taken from her, is always available to her at will (as far as most Rosharans know), is very hard to defend against, and no one would assume that a woman has one. Anything could be the case but I think that the Weeper is a lot like Adolin before he made the Unoathed possible: very important in the pre-True Desolation era, but simply eclipsed afterwards. I never expected the Weeper to be even a minor character, instead just a bit of worldbuilding flavor. I wouldn't mind being wrong though, she sounds like an interesting person no matter what else she might turn out to be.

This probably means that Shallan dies too. She's already a budding Radiant, which means Nale will almost certainly kill her. I'm not sure how expansive of a problem that is given the other consequences on the Shattered Plains (e.g., they don't cleanse Urithiru of Re-Shephir, but they may not ever have reached Urithiru anyways so what's the difference?). But she's a consequential person and someone that I think would have been caught up in events somehow. I have to agree that never having Kaladin probably dooms the anti-Odium forces. But I can also imagine Dalinar not changing enough to cause Taravangian to try having him assassintated (he might choose "the path of the warlord" instead), and therefore unify Alethkar by violence and domination. From there he pursues the pan-Rosharan conquest most people initially assume the coalition is attempting, largely unifying the continent (save except Shinovar, at least) while also weakening all of the nations through those conquests. Taravangian still cultivates Radiants and so has more than just mundane forces to command, but they are fewer and there is constant tension between their ideals and how Taravangian insists they operate. The Radiants aren't quite restored, but spren bonds give rise to Surgebinders operating more independently, maybe order-by-order but maybe even in smaller, more ideologically focused groups of likeminded individuals. Taravangian, as the only leader Roshar has, treats with Odium instead of Dalinar but without being the inheritor of Honor's power and so cannot necessarily release Odium. If he can release Odium after all, the bargain becomes trading Dalinar to lead Odium's armies in exchange for sparing humanity. Taravangian's age leads him to natural death not long after striking that deal, and his successor doesn't have any of the advantages Taravangian enjoyed and so fails to maintain many of the benefits of the situation Taravangian engineered. Cultivation flees and hopes to use her minor influence with Dalinar far into the future, or is splintered like Honor was. Either way Odium wipes out the spren of Honor and Cultivation, replacing them with his own spren and establishing hateful, vicious orders of Surgebinders. Odium's pressure against Rayse's will pushes Roshar into endless battles anyways, mainly to train soldiers for Odium's offworld conquests (much as the border conflicts among Alethi highprinces trained soldiers for the Shattered Plains). Retribution never rises, so the broader state of the Cosmere doesn't change very much and Rayse continues his campaign of wiping out the other Shards one at a time, now backed by loyal armies that prevent him from needing to expose himself to danger or needing to Invest other worlds very much. Rayse might even pick up the Dawnshard Rysn found, though Hoid would work to keep that away from him above all else and is probably clever enough to succeed. The far-future war tips very much in Roshar's favor due to a lesser sense of urgency and less time to prepare for the other Shards.

I was thinking more along the lines of "if I were an editor reviewing these books and had the cachet to persuade Sanderson to revise, what revisions would I suggest to improve the books?" and not "what can be done at this point to fix them", which I agree it is likely too late for. I'm curious, for you and anyone else who cares to comment, do you think that the character development and events related to it are similarly satisfying across the books? And, separately, do you think that the plot and major development are also similarly well done across the books? I ask because I do feel that the books accomplished some of the specific goals Sanderson set for them, but that isn't a big factor in how I evaluate them. And while I like the lore and bits of information that we consistently get across all of the books I can't help but feel that we could have gotten that information just as well even with radical changes to the story, so I don't give any individual book much credit just for including it. I persistently feel that there is a real distinction between the first two books and the latter three, and a lesser distinction between the first three and the last two and I'm interested in views that align or differ from that.

I'm not satisfied enough with the reliability of official statements to feel confident parsing it any more than I posted above, so the specific munitions depleted could well be just a subset. At the same time, the vaunted destructive potential of the U.S. military is also based at least in part on those munitions existing, so the effect might be similar either way. My main idea in bringing it up is that there probably aren't enough munitions available to flatten Roshar from the seas. As to the coercion, even when militaries have specifically bombed population centers it didn't have the effect you're suggesting. A terror bombing campaign is almost certainly impactful, but not a war-ender. It's the kind of thing that seems instinctively like it would be true, but people have thought this since Douhet in 1921 and it really hasn't worked out that way. That's a fair distinction to draw. I'm not even asserting that much, though. The bombing isn't going to be thorough enough for the invaders to simply win by doing it. That Rosharans have more resilience (due to magic) in infrastructure than real-world polities only makes it harder for destructive bombing to force an end to the conflict, and certainly less likely that they'd fold. But the destruction alone isn't going to limit what the Rosharans can do very much-- it's not even going to degrade their industry or materiel in a meaningful way.

For sure it's a different situation, but the core issue is the same: ability to blow up a bunch of stuff does not mean that you can necessarily accomplish any particular goals outside of that destruction. All the munitions the U.S. expended (a non-trivial portion of the total that was available) wasn't enough to destroy Iran's ability to launch counterattacks, still required staging through multiple bases, wasn't enough to force a change in government, and if it wasn't enough to destroy all military capabilities it follows that it also wouldn't be enough to destroy all population centers (certainly not enough to do both). Bombing population centers is also, historically, not all that helpful in winning a war or accomplishing war objectives besides maximum death, even if the invaders are capable of and willing to do it. The main point is that the bombs-and-missiles weren't enough to conquer a real-world opponent with inferior technology and far less wealth because blowing a bunch of stuff up is not, fundamentally, capable of that. The people who launched the most recent war with Iran assumed that the destruction they could unleash would cause it to fold very quickly and out of necessity, but it didn't.