The only option that isn't persuasive is the last one, as it is actually reasonably easy to provide an estimate of schizophrenia's heritability. (Compare identical twins reared apart to identical twins separated at birth. Any variation in schizophrenia incidence must be due to environmental differences.) Such studies indicate that schizophrenia is reasonably highly heritable.

Of course, this only deepens the paradox. If schizophrenia is so heritable, where are the genes? There are actually other options to add to the above list. If you're interested in the 'big picture' problem exemplified by the case of schizophrenia, try this accessible article from the Evolution and Medicine Review, entitled (helpfully enough), " Why are there so few genes of major effect on highly heritable disorders?"