Well, I think the problem we'd run into next is what is meant by "gameplay." My first point in this whole thread was saying the fiat/no-fiat divide isn't a useful razor, or rather it's an insufficiently useful razor. I don't generally find the critique of "processes of play" useful; it encourages a substitution of the process for the machine. It does not matter (except in quite specific aesthetic targets) whether we use dice or cards as a randomizer, or if a board game has hexes or squares to mark spaces. Those are technologies that are used to provide players with points of interaction within a game, and there may be a reason to use one or the other in the delivery of particular gameplay, the interesting decision laden here->there states I described earlier, but the mehcanisms aren't in and of themselves, those experiences.
Pages and ages ago, I presented "GM extrapolates from established facts" as such a technology precisely because it enabled specific kinds of gameplay loops. The problem that drives this whole discussion comes from substituting one kind of gameplay for another under the guise of adjusting/progressing the technology used to deliver gameplay. If we disagree on the deliverables, no amount of discussing the process of delivery gets us anywhere; this is what seems to have been the source of the "real mystery" disagreement.