Is it reasonable to assign P(X) = P(will_be_proven(X)) / (P(will_be_proven(X)) + P(will_be_disproven(X)))
No. For example, there exist statements that are unprovable but true. To assign a numerator of 0 (or epsilon) to all unprovable statements is wrong.
No. For example, there exist statements that are unprovable but true. To assign a numerator of 0 (or epsilon) to all unprovable statements is wrong.