If they are essentially procedural goals, where the motivation behind doing them is to find out whether or not they can be done, then yes, they are gaining knowledge about the shared fiction.
...
they don't raise any issues of real-world value/signficance other than curiosity as to how it unfolds.
So where we have a different understanding here is that the motivation behind doing these actions is not to find out whether or not they can be done. Rather, the setting's assumption is that these things can be done (and the reason they can be done is because the PCs
want to do it, and the players should be aware of this -- belief is power, your belief shapes the multiverse, belief is a key part of character creation because of it, etc.). The motivation to doing them is the same motivation that a PC in a "standard" D&D setting has for going out from the turnip farm and raiding the local mad wizard's ruin: they're the kind of person who responds to such a call to action (for whatever reason). The conflict comes, in large part, from the fact that the antagonists are doing the same.
I'm not entirely sure what "real world value/significance" brings to bear on the topic. I see the real-world value being the same real-world value that any game of D&D brings, which is engagement and entertainment. Again, maybe an example of the alternative would be useful for contrast, because I'm not sure I have a clear picture of what other issues of real-world value and signfiicance, aside from pleasure, result from any particular game of RPGs.
By fidelity I'm not meaning fidelity to external canon (though I'm sure for some groups that's very important) but fidelity to what's gone before.
So internal consistency = a limitation on player authority? Interesting take. I don't know that I'd disagree, but I also don't know that I'd enjoy a game that lacked that limitation (and thus lacked internal consistency).
You see this every time a poster talks about a "living, breathing world" or "the GM should play the NPCs in accordance with what they would sensibly do" or "I have a timeline that dictates how things will unfold if the PCs don't interfere".
I don't see the link between any of those things and a limitation on the player's ability to declare character actions and have the DMs resolve them.
The setting serves another purpose other than satisfying curiosity. I posted an example upthread with reference to Bladerunner.
Looks like I missed that reference, care to encapsulate?
In RPGing, then, the move from setting exploration as a priority to somethinge else (I'm hesitant to call it the opposite) is using story elements to generate real-world conflicts of value/matters of significance. Luke Crane gives a nice example in the BW rulebook:
If one of your relationships is your wife in the village . . . [then i]f you're hunting a vampyr, of course it's your wife who is his victim!
This example, taken literally, trades on a mechanic that BW has and D&D (other than perhaps some 5e backgrounds, depending on how some of the background features are interpreted) doesn't: players building relationships into their PCs as part of character build.
And by your criteria, that is different from witnessing the consequences of player actions in the setting? Because it really looks the same from here.
Heavy Setting Exploration:
Player (at some point in play): "To further the beliefs of my Xaositects, I rip a hole in the fabric of the multiverse."
DM (at some other point): "The Guvner looks at you with a mixture of awe and horror, recognizing you as the origin of the breach, as the glistening steel Inevitable strides forth with an earth-shaking INVALID."
Not Really Setting Exploration:
Player (at some point in chargen): "I have a strong connection to my wife in the village."
DM (at some other point): "The vampyr's trail leads right to your doorstep, and you realize, with a sinking sensation, that your wife is sleeping inside..."
Is the above an accurate representation of your case?
If so, can you tell me what you see as the significant difference?
I can see a few minor points of divergence. The latter situation has the player "taking an action" in character creation, rather than in play (in which case, simply adding more explicit goals to character creation would remove/relieve the issue). The proper nouns are a little different ("the multiverse" / "the village"; "Xaositects/Guvners/Inevitables" / "vampyrs"), but I don't imagine you'd begrudge settings different scales and NPC protagonist/antagonist groups. I might be getting your position wrong still, so I don't want to examine it too closely, but this is what I'm coming to understand.
In both cases, we have a player picking something that has some meaning to them (and in the PS example, demonstrating that meaning in play), and the DM then threatening that thing with people who would like to destroy it. In the PS example, it's an ideology, versus the BW example of a spouse about to be Damsel'd, but in both cases it's something that the PC is explicitly or implicitly willing to fight and die to preserve, protect, and advocate for.
But the same thing can be done in D&D on an informal basis. For instance, at the start of my 4e campaign I instructed each player to specify one object of PC loyalty. On this approach, to some extent the content of the fiction is already known, to the players as well as the GM. The GM's techniques of content-introduction become quite different (eg, as per Luke Crane, the victim will be the wife; or, in an example you've probably seen me use before, if the PC is a paladin of the Raven Queen then the "abandoned" tomb will contain skulking cultists of Orcus, or at least residue of their evil works).
...and if one of the PC's is a Xaositects, their antagonists will include Guvners or Mercykillers. And if one of the PC's is a Guvner, their antagonists will include some Xaositects or Bleakers...
When the focus of play shifts to resolving conflicts where the stakes have real-world heft other than "what happens?" or "will we win?", setting design has a particular role to play: it needs to contain the right material to generate these contracts. This is somewhat table dependent, of course - different people are moved by different things - but the staples on which I tend to fall back are things like loyalty vs betrayal/freedom/transformation; vengeance vs impersonal efficiency; what price to avoid death?; how ruthlessly will you treat your fellow party members to get what you want? (This last one is definitely "handle with care", at least in my experience.)
So, like, Harmonium (loyalty) vs. Fated (freedom), Mercykiller (vengeance) vs. Guvner (efficiency), Signer (what price to avoid the end of your world?), and, well, any game where a party of 5 is going to each individually have a stake in transforming reality?