What's the best way to answer this question in an rpg system? That is, is the role of the dm/rules in a roleplay game ever analogous to the role of umpire and rules in a kriegspiel wargame (free or not)? If not, then it would seem like your criticism of the realism of dragon flight extends to the trad approach.
In 5e, dragon flight 80' a round. Is that realistic? Is it valid in a gameist sort of way? Why does the fact that it's written in a book make it a better answer than the gm coming up with the answer, even at the table? I think the answer is that if you are playing 5e, you expect the rules to reference things in a way that somewhat guides or even constrains the ability of the dm to make those sort of rulings. Even if the dm is making it up as they go, there's an illusion that everything refers back to a pre-written stat block.
Similar for establishing the fiction. The fiction must be established one way or another; what means to doing so are appropriate? Should the dm prepare everything ahead of time, and does that prep make the establishment of the fiction at the table more legitimate (and if so, why?)? I think for all the emphasis put on dm trust in fkr, the more likely answer to any fiction-establishing elements would be more like, gm proposes, table confers, gm decides. High trust assumes that that conversation is not adversarial, and further that a non-adversarial social relationship among players and gm cannot be guaranteed by a set of rules or procedures.
I see the role of a book like the MM as being a substitute for personal prep. It substitutes for prep of the backstory, by providing a default one. (And in the context of D&D, despite the changes over the year, there is a clear trajectory from Gygax's MM with its devils and demons and mysterious underground drow and militaristic hobgoblins through to the 4e MM with its devils and demons and underdark-dwelling drow and militaristic hobgoblins; I can't comment on the 5e MM.)
It also answers mechanical questions.
I know - from hearsay, not experience - that there are D&D tables which put a high premium on the GM adhering to the mechanical details of the MM. I personally haven't had that experience; as a GM in D&D and RM (the two systems I've used which have extensive published "bestiaries") I've always adapted what the rulebooks hand me to serve my purposes when I want to.
But in the context of D&D (or RM) there are some other considerations that come into play, too. Movement rates often matter in combat resolution, and for that reason have to be kept within certain bounds that aren't necessarily "realistic" - eg in AD&D the default human movement rate is 12" and really fast flyers go up to 48" (eg a giant eagle' even an air elemental is only 36"). Now in real life an eagle can comfortably fly at 50+ mph (as per
the ever-handy wikipedia; I remember 20+ years ago tackling this question during a session - when a PC changed shape into a goose and wanted to fly from A to B - by pulling an encyclopaedia from someone's shelf). Which is quite a bit more than 4 times human walking speed.
But do we want shapechangers, familiars etc to be able to move so fast? And how about the time to stop and start, which is a factor in a lot of D&D combat movement? For a range of both balance and common-sense reasons, we compromise! I would expect the authors/designers of a MM to have regard to these sorts of concerns in settling on their movement rates; and as a referee doing my own designing I would pay attention to what they have done in that respect. (A side remark: 4e D&D tried to tackle this issue, in part at least, with the concept of
overland flight speed, but that didn't really survive beyond the first MM and was never systematically operationalised. Not every design innovation is a success!)
Now here's another example, more recent: in my Classic Traveller game the PCs had called in support from the air force of the world they were on (one of the PCs was an ex-army colonel and had good connections with the world's military). It was already established that the world's atmosphere was toxic and corrosive, and that had led me to the view that the airforce would be prop-based rather than jet-based (because (i) it wasn't clear that the atmosphere would support combustion as a jet needs it, and (ii) I thought the corrosive atmosphere would damage the jets more seriously than propellers). So I explained this to the players, and they agreed that made sense. And then I suggested a flight speed for the air force planes as around 300 kph, and no one dissented from that (including the military history buff). But a subsequent Google revealed that to be woefully slow: even the passenger Fokker Friendship I flew on in the 1980s has a speed of around 500 kph. And given that this is a game of science fiction adventure in the far future (the sub-heading on my Traveller rulebooks), realism matters!
Because I don't use map-and-key resolution, my error with the flight speed didn't matter - it was just some colour to give content and context to a framing/pacing decision. But in a map-and-key game, where the fate of the PCs might depend on how long the air force planes took to arrive, it would be more serious. I wouldn't blame the players for being irritated by it.
