Verifying rationality is possibly the single largest problem for those desiring to create methods of systematically training for increased epistemic and instrumental rationality - how to verify that the training actually worked. Very Awful Things happen to people who set out to build training methods and schools without strong means of verification - two hideous historical examples being modern-day practice of martial arts (once the teachers are no longer fighting duels to the death) and the proliferating "schools" of psychotherapy (in the entire absence of any experimental evidence that one school worked better than another, and indeed, in the presence of experimental evidence that the schools did not have any discernible difference in effectiveness).