Mathematicians are definitely right that the value of finding and solving puzzles is the generation of new puzzles. And I think historically it's clear that incorrect findings eventually get sorted out. So the incentives should probably be put on the former over the latter.
