Well, I suspect its assumed the latter represent some in-fiction process mostly, whereas metacurrency is explicitly (except in some odd cases like TORG) just what it says on the tin.
Even putting TORG to one side, I don't think this is right.
I believe that
@FrogReaver and probably
@Crimson Longinus regard a Circles or Wise check in Burning Wheel as a form of "metagame" "narrative control (although it is based on a check rather than a currency). I believe that they would see a player's use of a Special Effect (via a Storyteller Certificate) in Prince Valiant as a form of "metacurrency".
But in all these cases the process at the table - of making the check, or cashing in the certificate - represents an in-fiction process. A Circles check corresponds to looking around for and/or hoping to meet a person. A Wises check corresponds to trying to remember the details of something. And I quoted the relevant rules about Special Effects in Prince Valiant upthread (from p 44 of the rulebook): "The Storyteller must create a reasonable explanation for the way in which the Effect takes place, in terms of the current situation."
So whatever it is that sets the conventionally-received boundary between "metacurrency" and other stuff, I don't think that it is about in-fiction processes. Rather, as you said upthread:
where does the no-die-roll, straight-declarative authority stop? I think some of the functions of some forms of metacurrency are to keep a largely traditional separation here while permitting limited ability for the player to put his oar in in areas where most older trad games wouldn't.
Those traditional separations/boundaries aren't explained by
representation of in-fiction processes. In addition to the examples I've given, which
do represent such processes but I think are controversial in a traditional context, it's widely accepted that a player can specify stuff in their PC's orbit (names of parents; colour of cloak at the start of the campaign; etc) without those decisions having to represent an in-fiction process (eg the PC didn't name their parents).
This is why I prefer an analysis in terms of
backstory and
situation, which I think actually does track what is going on pretty closely:
* The GM has primary authority over backstory - some minor details within the PC's orbit are the exception (as I said just above). It's easy to find debates about where this boundary lies - everything from GMs policing race and class choices, to people arguing about whether random family generation charts are a good or bad thing.
* The GM has primary authority over framing situations - but there are significant differences here in the use of eg random reaction tables, random weather tables, etc, which can be seen as imposing constraints (ranging from the minor to the very significant) on the GM's framing.
* The big point of difference is
the basic principle that guides the framing of situations:
* Is this largely fiat, with backstory then adjusted (behind-the-scenes) to support it? It seems pretty clear that this is what happened in
@hawkeyefan's Folk Hero episode of play - that is, the GM had a scene he wanted to present, and he manipulated backstory (about how the duke's soldiers discovered the PCs) in order to support that framing. But the more this is done, the more the game drifts towards a "situation first" game of the "living novel" sort, and the role of backstory becomes less and less important as an input, and more something that the GM just develops as needed to support the series of fiated situations.
* Is this highly constrained by pre-authored backstory, whether map-and-key or the "evolving" backstory of a "living sandbox"? This is what Gygax and Moldvay present as the norm - they have no objection to the GM writing interesting stuff, but that is presented as something to be done as part of prep, in designing the dungeon, and not really as part of
play (an exception, which sits a little uneasily with the general tenor of his advice, is Gygax's suggestion that the GM might manipulate dice rolls to support more interesting framing, on pp 9 and 110 of his DMG). Just as it's possible to write an interesting dungeon during prep, and rely on the prep to generate interesting situations in play, so I think the same can be done for a "living sandbox", as
@Campbell has outlined above.
* What is probably fairly common, I think, is a mix of the above two approaches - ie the GM does prep, and in principle treats it as constraining on situation-framing, but adjusts or departs from the prep when that seems desirable in order to fiat a situation.
One thing that I find interesting about response to
@hawkeyefan's example of the escaping hag is the relative uniformity of critical response:
one point of frustration I experienced as a player.
<snip>
I haven't yet had a chance to discuss that specific point of play with him, but I plan on it. I found it to be pretty frustrating.
The first I would have personally found irksome.
At minimum, this seems like a hella clumsy. Though it also kinda depends on how distances are normally handled (was there a battlemap?) If distances are always handled somewhat abstractly, then this might be more understandable. Though coming up bunch of reasons why you couldn't attack seems like BS either way. Better hope that if your characters ever need to flee the GM handles that with equal generosity!
Only
@Ovinomancer seemed comfortable with it:
These both seem run-of-the-mill uses of GM Force. Seems pretty typical of 5e play to me.
Now if one takes the play loop of 5e D&D literally and as admitting of no further qualifications, then this episode fitted it: the GM described the situation (fleeing hag), hawkeyefan declared his PC's action (ranger tries to shoot the hag) and the GM narrates what happens next (the shot is impossible and so no roll is required).
We can quibble at the edges - it seems that the GM may not have described the outcome of a shot ("you miss" with no roll called for) and rather have described the ranger's state of mind ("you realise shooting the hag is not possible") but I don't think that's the crux of the issue (in a sense, that is doing hawkeyefan a favour by helping to conserve ammunition). The crux of the issue, as I see it, is that most RPGers
don't regard the GM's authority to narrate what happens next as unconstrained: in this case, rules for range (as found in the weapon rules, the feat rules, etc) and for spotting enemies (as found in the Perception rules, the spell rules, etc), as well as a general sense of "fair play", are expected to constrain that authority.
I think this shows that, whereas in traditional RPGing there is a diversity in understandings of how scenes should be framed, there is less (not no) diversity when it comes to the action resolution process. I think most RPGers, even traditional RPGers, are suspicious of the GM just suspending the standard action resolution rules.
And I'm prepared to conjecture even a little bit further: I think that if hawkeyefan's GM had used his authority over backstory and situation to conjure up a rules-legitimate way for the hag to escape (eg a newly-authored powerful friend teleports her out), that might be seen as a bit dubious by many traditional RPGers, going as it would not just to initial framing (as in the Folk Hero case) but directly to the resolution of a conflict-in-progress.
A player can be very sceptical of a certain sort of meta-currency but still have these sorts of reasonably clear views about what counts as acceptable exercise by the GM of their authority over backstory and over situation.
EDIT:
Oh, like hitpoints! And AC points! And Death Saves!
ETA: the opposition to metacurrency always seems so much special pleading to me.
That's why I don't really regard the issue being one of either "metacurrency" or "in-fiction processes". It's about "domains" (for lack of a better word) of the fiction, and who has what sort of authority over those.