OK, then let's go back to the OP - it asks, what is this allocation of responsibility to the GM for?Why does the DM get to decide? Because it's the DM's job to set things in place within the dungeon and gameworld - this is part of what makes a DM's role different from that of a player.
In Cortex+, the roll to find the map is an opposed check (against the Doom Pool - if the player succeeds, a Map To XYZ asset is created; if the player fails, then maybe the Doom Pool steps up, or the PC suffers emotional stress in frustation at not finding the map, or whatever other consequence flows from the mechanics of resolution plus the imagnation of the GM). In BW, it is what you call a "passive check" but against a difficulty set by the GM in accordance with the skill descriptions; but that mechanical difference doesn't mean that the GM gets to make the passive check fail automatically just because s/he thought it would be better for the map to be somewhere other than the study.
In other words, (i) there is not only one model for RPG mechanics, and (ii) even when the mechanics are similar (both D&D and BW use checks against a difficulty), that doesn't tell us why it is the GM's job to do the stuff you say.
To be clear: I'm not asserting that there is no answer to the question. But answers that don't take account of the range of ways RPGing works will (necessarily) be incomplete. I mean, obviously setting provides depth - but it doesn't have to be GM authored to do that (witness the various examples I've posted upthread). So a more complete answer adds information eg [MENTION=284]Caliban[/MENTION] says that many players don't want to contribute to establishing the backstory, so someone else has to do it; [MENTION=59082]Mercurius[/MENTION] says that he wants the GM to tell him the backstory as part of his process of immersion (to me that seems very similar to being told a story by the GM - I think Mercurius queries that characterisation, but from my point of view I'm still working out why, and also why it's considered pejorative - I went to the pictures recently, and had a story told to me, and that doesn't make me feel offended).
[MENTION=23935]Nagol[/MENTION] gave some different reasons: GM worldbuilding establishes levers/tools for the players. It makes sense that someone else has to do this, in that being able to just deem your own tools into existence seems a bit cheat-y. To me, that speaks to a style of play much closer to classic dungeoneering, though mabye Nagol would not agree with that.
Why? And which game are you talking about?Also, the very term "action resolution" is here a bit misleading. Yes a PC has declared an action, and that action gets resolved...but the resolution of that action only applies to the PC and her immediate surrounds, not to anything static within the rest of the game world.
In Classic Traveller (1977 version), the rules set throws required on a player's Streetwise check for a PC to find a shady official willing to sell permits/licences at a good price. That is an action resolution that is not confined to the PC and his/her immediate surrounds. If successful, it estblishes that said official exists and is willing (everything else being equal) to sell permits on the side. (Not all of Classic Traveller is like that - the rules for finding the Psionics Institute specify a GM-side roll to establish its presence on a world; then a player-side roll to find it, which can succeed only if the GM's roll turned out right. The rules don't discuss why pisonics is handled differently from Streetwise, but I think the idea is that the GM is expected to gatekeep psionics to a high degree, whereas finding officials to sell permits is a central part of play.)
In AD&D, a paladin can call for a warhorse which then obliged the GM (per Gygax's DMG) to create a whole backstory and "side quest" around the lcoation of that horse, and the mission the paladin has to complete in order to win it.
There is no in-principle reason why finding a map can't be treated in the same way as the above examples; so if it is not being treated that way, why? (I've offered a conjecture as to the different treatment of psionics in Traveller, just above; but why does the GM need to gatekeep the location of the McGuffin?)
As I've explained, I don't find these reasons very convincing. There is nothing unrealistic about the map being in the study. (It's not like finding beam weapons in the Duke's toilet!) If no prior contrary backstory has been established, then it's not inconsistent with anything.For consistency, realism, and believability it works better the other way around, where the action resolutions are bound by the constructed world / setting / dungeon.
If the GM hadn't decided, in advance, where the map was - maybe she hadn't got around to it; maybe she left all her notes on the train by accident on the way to the session, and so is doing her best to remember them but has forgotten the bit about the map - then something would have to be made up on the spot. That happens all the time in RPGing. It doen't make games less realistic or believable.
Instead of the GM just making it up, or rolling some dice secretly, it can be made consequent on the player's roll - as Traveller does for dodgy officials and Streetwise checks. That doesn't make the Traveller gameworld unrealistic or unbelievable.