Okay, you and
@AbdulAlhazred seem to have (slightly??) different takes on what No Myth entails and I've seen plenty of proponents on these very boards, some liking your post, claim that you are not supposed to have things pre-made. But given that your premise above is correct, how is this enforced by the game and how can the players be sure it is not being subverted by the GM without their knowledge.
This seems to me to come back to a fundamental point about action resolution.
"No myth" play, as
@loverdrive says, is compatible with thinking of things in advance.
What "no myth" play excludes is
using that stuff that was thought of in advance as a basis for deciding, in advance, how things go in the fiction
independent of the action resolution mechanics.
Now, this point has implications. If the action resolution mechanics
can't be applied except by reference to stuff that has been thought of in advance, then the RPG in question will not work well for no-myth play.
The most obvious case I can think of that raises this issue is the resolution of interaction with NPCs. If there is no social resolution system (eg the Classic Traveller reaction table), and the GM is expected to resolve how NPCs respond to the PCs by reference to their already-known personalities etc, then no myth probably can't work.
That said, another distinction is relevant here: between framing, and secret backstory. The following could be an example of no myth: the GM narrates the NPC saying "I don't really want to <do such-and-such>, but I would love for you to give me <this thing>". That's framing, and the player can respond to that: just as if the GM frames a scene involving a pit and a wall beyond it, and the player responds with some action declaration involving a rope, grapple and acrobatics.
On the other hand, this is not an example of no myth: the GM makes a note to themself that NPC really doesn't want to <do such-and-such, but would love to be given <this thing>. And now the success or failure of the players' action declarations is adjudicated by reference to this secret backstory.
In passing, this also shows that nothing about "no myth" play precludes what has been described in this thread as "simulationist" reasoning. All it does is put additional limits on what the GM can extrapolate, and how they present that to the players.
Here's an example... A manor is established in play that is owned by a benevolent mayor who really has his town's best interests at heart. The players are invited to the manor as acting diplomats for the town to discuss the rising tensions in the political state with a group of representatives from a secondary village to the north. There is a feast, introductions and the true negotiations are declared to start tomorrow. The players are all ready for a game of political intrigue, negotiation and closed door deals... I however as GM want to run a dungeon crawl I wrote up last night full of ghouls, undead and body horror. So I have decided that under the manor is a dungeon secretly built by his son (a secret cultist) who, when the players sleep at the house for the night, drugs them and dumps them into the ghoul-infested, haunted catacombs full of body-mutatring fungi. Now going strictly by what you said this doesn't violate No-Myth.
This is clearly not a no-myth game. You have all this secret backstory, and you are making hard moves by reference to it, bypassing any action resolution framework.