I'm focusing not so much on the +2 dice as on the fact that a Stunt both enables the impossible and completely negates the penalty/Difficulty Increase for doing the impossible. That seems very logic-rejecting to me. Both seems to be based around the guideline of 'if it looks awesome/etc., then it is okay to ignore the normal consequences in favour of awesomeness'. This is why many alternate history settings tend to have zeppelins in modern times/alongside modern tech, actually (TimeShift, Red Alert 2 & 3, IIRC Gernsback, Sky Captain and the World of Tomorrow, Iron Sky, Equilibrium &c.).
Uh, no. Again, we know what the stunt rule is for, I just told you. It's there to get players invested in a scene and actively describing their actions. For incentives, it provides a +2 dice bonus and 4 motes or 1 willpower, and also protects you from the GM being a vindictive little prick and screwing you over for attempting to do something interesting, so there are no downsides and plenty of upsides, describe away, do cool wuxia stunts, copy your favourite chanbara film, whatever.
Why in the world would you take that and extrapolate it into a blanket assumption that nothing in the setting is supposed to make sense, from "what ships work in what weather conditions" up to "humans act like humans"?
'Assumed to be true; the GM did not' is precisely why I'm leery of undocumented headcanon changes. They make complex interactions prone to a chance of failure through no fault on the player's part.
Yes, well, "every group has their own canon list and set of books that don't exist" is not an abstract proposal, it's how the game is actually played. Because global consensus does not exist and it is impossible to do RAW, unfortunately. Therefore, if you want to try popping into just about any Exalted discussion on the interwebs and start insisting that everyone stick purely to exact published canon, you're likely not gonna have a good time.
Outright contradiction between explicit parts of the setting definitely cause problems of a similar type on their own. Hoever, an explicit bit of the setting contradicting an implicit part of the setting (e.g. it's AFAIK not stated anywhere that all paths from Malfeas to Creation take one day, though it does say that going through Cecelyne does; fast path contradict the implied but not the explicit bit) don't seem to cause this type of problem. Instead, they produce a mystery that one can explore, finding hidden reasons for an unexpected phenomenon(-na).
Five days, and we know the reason why because Borgstrom actually came out and talked about it. Some random freelancer writing something that ignores it is a case of... some random freelancer not knowing what he was doing. Because people like the five-day lag and agree with the reasoning behind it.
You can certainly make something interesting out of it in your own game by introducing an instant Hell Portal and nobody will stop you, but trying to get others to go along with it in the form of "okay, that in fact
should exist, screw the five day lag, the freelancer was absolutely right!", that's... probably not gonna happen.
Now, SMAs in Monk are . . . yeah.
And people reject those because they're contradictory with, well, just about everything else. Are you going to quixotically insist that everyone talking to you treats those bits of setting information as definitionally good by virtue of being published too? No, yes? I gather you, too, agree with banning Obsidian Shards of Infinity. What makes this particular contradiction unacceptable while all others seem to be something you, hm, actively like?
Well, yeah, after researching an explicit list regarding what is untrue. Of course, I can't mind-read those lists, and it takes very long to actually read a bit-by-bit list for the whole setting even if one exists for a given poster.
Except it doesn't matter in the slightest if you're talking to some random on the internet, it only matters when you're playing a game, at which point your group sets its own list and you have local consensus. If you're looking for some kind of globally agreed on list to stick to when interacting with other people online, I can only point out, again, that there isn't one, and trying to insist on "everything published exactly as written" as said list will not work.
Example of a discussion I remember seeing a while ago, truncated and paraphrased.
Poster A: So guys, how do I best use Violet Bier of Sorrows Style?
Poster B: Ha ha ha, no such thing exists!
Poster C: Nonsense, Violet Bier of Sorrows Style is perfectly balanced, and Poster A should teach Blade of the Battle Maiden to his entire circle.
Poster D: Here's a link to a rewritten version that isn't broken!
Poster B: That doesn't fix all the problems!
Poster C: It doesn't need fixing!
Etc, etc. The scenario where all four posters stop talking about whether Violet Bier of Sorrows should exist and instead start discussing the merits of the published material as written is a nonstarter, it basically won't happen.
Uh, I did. But we seem to have a certain difference of premises. Notably, you expect a total newbie GM going all-out. I'm expecting a GM who has some experience with RPG systems in general, and who is not aiming to TPK.
If one pits two equal opponents in a life-and-death combat, then there is 50% chance that one of them will die. When one pits two equal groups, things get complicated, but the chance of at least one side losing one member is still high (50% unless there are some weird mechanics that allow one to tone down life-and-death'ness in group combats). So I don't expect confrontations with fully equal opposition to be routine and/or don't expect confrontations with high-power opposition to be life-and-death.
You seem to expect that all NPCs are built as optimal powergamers, and go all-out. The consensus at the table is that real people (PC and NPC alike) normally aren't optimal.
You have posted displays of the death spiral as a dangerous thing. I'm saying that it is not necessarily the case that the path between the beginning of the death spiral is a white room, and that some sort of unforeseen factor may save the PCs in an over-the-top Indiana-Jonesey manner.
Or maybe I'm misunderstanding and you're pointing out something different?
/facepalm