Individual examples do not contradict the general statement. Galois was dead at an age when I wasn't even at university. Abel a bit later (so avoid groups if you want to have a long life), same with Ramanujan (which incidentally may be a factor in Hardy's comment). And so on, just as singular at the first glance. As a mathematician, however, I continue to argue Hardy's point, both for the present and for the past as a general and observable phenomenon.

And the number of books as a measure of quality, really? I think that view is skewed by today's “publish and perish” environment (nothing against Euler).