I've got increasingly distrustful of simple dichotomies for RPGs. While these can be of some value in some or even most games, they aren't universal, and their usefulness is a relative thing.
Agreed. I think the whole idea of a gamism/simulationism axis (or a gamism/narrativism/simulationism triad, for that matter) is fundamentally flawed. It's not a useful way to approach RPG design.
To me, the goal of D&D (on the player side) is to provide an exciting experience for the players, by allowing them to imagine themselves as their PCs and test their (player) skills against the challenges the DM puts in front of them. The challenges may be strategic, to be met with players' tactical skills; roleplaying, to be met with players' roleplaying skills; or, in most cases, a mix of both tailored to the particular group's tastes.
A certain amount of verisimilitude is required for this to work. Verisimilitude has nothing to do with accurate simulation, however. As I've said in several other threads, it's about making the fictional world believable in the moment. The rules are like the set in a play. They don't have to hold up under close examination. If you walk up on stage and look closely at the set, the fakery is obvious. And after the play, you might have a "fridge logic" moment where you realize some details didn't add up. That's okay. It'd be nice if it didn't happen, but it's no big deal if it does.
But the set
absolutely must not distract the audience from the play in progress. If the audience is questioning the set, you don't get to stop the play and explain how really, it all makes sense and fits together. The moment they started paying attention to the set instead of the actors, the set had failed. If,
during play, you have a moment when people look at the rules and go, "Uh, what?" the rules have failed. It doesn't matter how many explanations you've got. The fact that you have to explain at all means you've lost.
On the flip side, as I said, D&D needs to let the players test their skills both strategic and roleplaying against challenges posed by the DM. Good mechanics support that by helping the DM construct a variety of engaging challenges, and providing a "common ground" of principles by which those challenges will run. That common ground ensures that the players have some basic tools to address those challenges, tools whose workings they understand.
Strategic depth and verisimilitude are not opposite ends of a spectrum. They are goals to be pursued together. There may sometimes be tradeoffs between those goals, but I think those cases are far fewer than people believe, and focusing on the tradeoff encourages the designers to settle for rules that aren't as good as they could be. The thinking is, "Sure, this rule tends to break verisimilitude, but it's good for strategic depth and we want to nudge things in the gamism direction a bit, so we'll go with it." It should be, "Well, this rule is good for strategic depth but it breaks verisimilitude. How can we improve verisimilitude while keeping strategic depth?"