I don't know, at this point it just feels like it becomes a catch all, if Tolkien's statement is in support of simulationism, then anything can be classified as simulationist, including Permerton's examples of his PBtA play.
Reconsidering Tolkien as describing a
technique and its effects, then it can be seen as having particular value toward simulationist purposes, without having to see that in an exclusive light.
So Tolkien is I think advocating consistency, but not just with whatever happened before, but with world laws. What is meant by "laws"? Well, that is left rather undefined. Generally speaking, I think laws constrain not just the case at hand, but go on to constrain future cases so that if we know the world-law, we know what sorts of things can or can't happen in our game world. They give an overarching coherence: not just that elves fly because they fly, but elves fly because of some world-law that accounts for flight in every case that falls under it.
Maybe a world in which elves flies feels a bit ridiculous to me? I don't take Tolkien to be saying that all world-laws are created equally. They're necessary (or it might be better to say extremely valuable) but that needn't make me think they are sufficient.
E.g. that lightning strike one that seemed to cause debate whether simulationist or narrativist - if in the game it hadn't previously been shown that lightning strikes (whether magically triggered or not) can't destroy half a house, then there is no law as such being broken, just a new law being introduced, so it is simulationist.
I take this to be where Tolkien's observation of the effects of consistency with world-laws matters. Certainly I can "discover" my world-laws as I go (and probably can't avoid that to some extent). But what happens at that moment of discovery, when there is still no world-law in place? Then I can't rely (in that instance) on consistency with the not-yet-formed world-law, producing the complaints raised in the thread.
This again argues for seeing Tolkien as describing a technique and its effects. As
@pemerton reminds us - simulationism is an agenda or purpose, not the internal logic or structure of the setting. System continues to matter: Tolkien's technique is one that matters to simulationism... it helps achieve it. As a technique, it can be implemented poorly or well, it might not be needed in all cases, and other techniques might aid and abet it. None of that rules it out as a technique of simulationism. And this is like saying that alliteration is a technique of poetry: that doesn't rule out its use in prose.
Whereas I tend to view simulationist as having some sort of goal to measure by - e.g. if Tolkien's world is a purely fantasy setting, where normal physics don't apply, then the non-magical flying elf is fine. If Tolkien's world is supposed to be a historical view of earth and creatures are supposed to act like earth creatures except where magical causes are in play, then the flying elf isn't fine.
Elsewhere I have argued that a basic principle of simulationism is that world facts are established in view of some reference. Tolkien is suggesting that world-laws can supply such a reference. Such world-laws will be incomplete. Rather there will be some set of world-laws that make the authored world distinct (such as those arising from theological commitments) while much else will use the real-world and perhaps some set of pre-existing texts as references. So in the absence of a world-law saying wingless humanoids can fly, we'll rely on norms established in view of our real-world. But perhaps if we had a canon of Scandinavian myths in which elves normally flew, we'd be comfortable with their flying.
Thus it comes down to - if you think Tolkien's statement is in support of simulation, then what is the simulation of? And the only context in the sentence suggests it is simulating a world where some all powerful authority figure can enact whatever it wants - and I'm trying to make this separate to the author, as an author writing anything he wants isn't simulating something, he would have some goal in mind that we can measure against if simulating or not. Thus the only goal that seemed to be able to measure towards to say his statement is in support of simulating, is that he is simulating a world where an all powerful figure (perhaps Illuvatar) can make anything he wants happen, bound only (and loosely it appears) by consistency.
Game worlds are authored either directly or indirectly. That needn't be by some all powerful authority figure. Were there such a figure, world-laws could do a good job of constraining what they say. Having established such laws, they ought not to go on to say anything that breaches them. Or if they do, then they are setting aside Tolkien's technique.
I can imagine an RPG where everything is decided by dice rolls referring to tables, but the tables are detailed enough to make it a good simulation of a particular setting / time period, vs something that is purely DM authored, and DM is consistent to anything that has happened before but otherwise just makes it all up on the fly and can come up with many weird and wonderful occurrences, and wouldn't be simulating anything.
I think you implicitly answer your own doubt there. You're drawing attention to the incompleteness of the world-laws. Can you point to any TTRPG game text that says
everything necessary and sufficient to cause participants to say things that make sense?
That doesn't doom the project. I just say that the world is like our own except that it has such-and-such set of world-laws, and except that anything implied by some reference canonical texts is taken prima-facie to be factual. I necessarily - in all TTRPGs - rely on a bunch of pre-existing capabilities and norms.
One of the hestitations folk have with simulationism is the idea of "simulating" a reality that comes into existence through ludic authorship. We're finding out what that reality is like. All this says is that "simulationism" as a label in TTRPG has a different meaning than it has as a label in other contexts: Edwards pointed that out years ago.