On the contrary: you are ignoring the role of epistemic uncertainty, which makes the metaphysical aspect of the situation secondary.
The character does not know what the runes say; but they have a conjecture and a hope. For them, epistemically, it is possible that the runes reveal a way out.
The player does not know what the runes say, but they also have a hope. For them, epistemically, it is possible that the runes reveal a way out.
What resolves the epistemic uncertainty for the character is reading the runes. What resolves the epistemic uncertainty for the player is a resolution process that tells them what happens when the PC tries to read the runes. The fact that the underlying metaphysics are different in each case - in the fiction, the fact about what the runes say becomes known to the PC; in the play of the game, the fact about what the runes say is established via the resolution procedure - is irrelevant. It doesn't change the epistemic situation of either the player or the character.
In a GM-authored-backstory game, nothing changes for the player in respect of their epistemic uncertainty: they don't know, but they hope. And so it remains epistemically possible for them that the runes reveal a way out. What does change is that the uncertainty is resolved not solely by a dice roll, but rather by GM decision-making. For some RPGers this may make for better play - for instance, it permits the game to be a puzzle-solving one - but it doesn't change the basic epistemic situation, nor the relationship between the epistemic and metaphysical situations.