Fundamentally what we're talking about is how situation and setting is determined by the play group. How do we acknowledge that in one case it is determined by a GM ahead of time writing some stuff down / defining things / thinking about how it must be while in the other case it is being determined in the moment based on what speaks to the interests of the player characters and the ongoing narrative as experienced so far? How do we acknowledge this difference without treating the second as if it somehow involved more artifice than the first?
I meant it would certainly help if we stopped referring to a complete, fictional setting as an illusion that one will eventually learn to see through. That's usually the point in these conversations things break down. It's difficult to empathize when the other party views the basis of the game you're playing as a childish delusion that can and should yield to better technologies.
I’m distinguishing between the GM’s ability to exercise judgement and the computer’s rote processing of its instructions. The “map and key” is input into the GM’s decision-making, but it is not the limit of the decision making. All the computer can do is follow instructions by rote. While they are both “data”, how they are used and their users’ relationships to them differ.
That’s why I call it a stretch. It’s just a tool. Some implementations are better, and a lot are worse. Trying to construe it as being more than that muddles things and risks saddling the tool with a bunch of extra baggage. (And it’s already bad enough that high quality keys are not as appreciated as they should be.)
So this distinction is particularly important, and does suggest that we should find a more specific term if necessary. The primary reason that a GM is a more compelling technology than a computer for this kind of play, is that they can be asked to go back and fill in whatever details they missed earlier. You can ask the GM to step back from adjudicating the game and/or making NPC decisions, and return to the role of worldbuilder to add additional details to the aforementioned map and key if they become relevant, something a computer cannot do outside of tightly controlled circumstances.
The whole reason you'd want to do that, is because it means you can a much broader set of actions to declare than a videogame can handle, and thus a much broader set of interesting decisions to grapple with and optimization cases to sort out.
I was trying to explain something about board games in a different context lately, which I think is relevant to understanding the agreements that this kind of play hinges on. None of my friends that I go and play cutthroat rounds of Indonesia or Pipeline with care about winning. We occasional track scores and what have you, but mostly because we're interested in the differentials between final point values, and how they differ as we play more rounds of the same game, but fundamentally none of us really cares who won at the end of any given round.
We all mutually agree that we will strive to win, because that agreement allows an interest set of board states to unfurl, and decisions to be made for us to chew on, which are the actually enjoyable part of the experience. This means, however, that we spend a lot of time giving each other advice, or discussing two or three options, and so on. Any move that results in an obvious advantage for a competitor is quickly pointed out by that competitor and usually taken back, unless we're playing in a specific kind of game where mistakes are considered essential, like say, Guards of Atlantis. Say, I could have taken 2 different sets of 3 pipe tiles with the same number of blue pipes but one of them has the orange pipes my opponent clearly needs, if I don't take that set, that opponent will likely point to them and say "you should probably take these, that will hurt me worse, unless you're doing something I'm not seeing."
So, the point I'm looking for in my RPG play is to produce that same decision making, ideally on a much broader scale than a board game can accomplish. I have yet to find any better way to do it, than having one person attempt to faithfully simulate a fictional reality that can relied on as a board state to act against.