I disagree... by this definition realism does in fact constitute simulationism
I don't see how you're getting that out of my post.
What I said was that, as far as the story of LotR, or the Marvel Universe is concerned, it
doesn't matter whether or how Glorfindel I and Glordindel II are related, nor what the relative strengths of The Thing and The Hulk are,
unless this matters for the story. Whereas in simulationist game play, these are the sorts of questions that become pressing. Because simulationist game play prioritises exploration of the fictional world.
This has nothing to do with realism. It's about internal consistency.
In purist-for-system simulationism, the focus is consistency of ingame causal logic. Rolemaster, Runequest, Classic Traveller, and big chunks (but not all) of 3E are built to with this priority in mind. Realism is a factor here, although not the only factor, because the real world is our main inspiration for what counts as coherent causation.
In high concept simulationism, the focus is on consistency and coherence of the relevant genre tropes and story elements. Call of Cthulhu and Pendragon are great examples of this.
Malory, Cretien de Troyes etc, in writing their Arthurian tales, don't have to settle the question of whether Percival or Galahad is the more pious, the more charitable, the more noble of bearing, etc. Once you start statting them up in Pendragon, however, those sorts of issues have to be settled. Likewise in Cthulhu - what is more sanity blasting, a horde of deep ones, or a shoggoth?
It's not about realism - in the real world, there are no sanity blasting aliens, and "nobility of bearing" (at least in Australia, and I imagine America) is a concept with about as much applicability as those that appertain to duling and to honour. It's about consistency and coherence among the story elements.
I believe you can in fact simulate S&S and high fantasy stories
Of course you can try to. Balesir's point, as I understand it, is that it is non-trivial to do so. It is particularly non-trivial if what you want to
achieve is genre (high-concept) simulation, but the
means you use are purist-for-system oriented mechanics.
Cthulhu provides a good example here, I think. As with many pulp or pulp-ish stories (eg Indiana Jones, Tintin), for CoC to work we have to completely ignore economic and institutional questions like "Where do these guys get their money?", "What reasearch institution is paying for all their non-teaching time?", "Why do they never have to meet deadlines for the submission of copy even though it says 'journalist' at the top of the character sheet?" etc. So good high concept design, for a game like Cthulhu, will push these quetions to one side, and shift the focus of the mechanics, and the focus of play, somewhere else. Conversely, if you start a session of Cthulhu with the PC professor being interrogated by his dean as to why he's been remiss in supervising his grad students, and with the PC journalist being sacked for repeatedly missing deadlines, the game is probably not going to head in the genre-appropriate direction.
As I understand [MENTION=27160]Balesir[/MENTION]'s point, it is that classic D&D,
at one and the same time, wants to achieve the genre feel of sword-and-sorcery, but wants to use purist-for-system mechanics (demographic models as seen in various DMGs; serious attention to buying and selling of goods and services, including magic goods and services; etc) to get there. And
this is what makes it hard to run simulationist D&D. For it to work, there's stuff you have to turn a blind eye to - just as CoC depends on turning a blind eye to the realities of employment as an intellectual or a private detective or a police officer - but the mechanics keep pulling the focus of play back onto those very things.
Now of course one can just ignore the mechanics in question - but now we're starting to talk about how easy or hard it is to drift D&D towards simulationist play, which is a conversation that already concedes a good chunk of Balesir's claim (at least as I understand it).