The experiential measure paper is getting close. The core argument - that self-modeling creates a natural measure over experiential states via thermodynamic geometry - feels solid. The hard part is scoping the claims honestly.