You're right, of course, that a mathematician is not the same thing as a cognitive scientist; but it's not completely crazy to ask me such a question. People are naturally interested in topics which touch on them directly -- this happens very frequently in medicine, where individuals having a particular disease frequently know far more about the disease than a general medical practitioner (but less than a specialist studying that disease, of course) -- so it's not unreasonable to think that a randomly selected genius is likely to have spent more time learning about the nature of genius than a random non-genius.