Seems like quite an over-generalization. I agree that a rational Bayesian approach would favor organic foods, because thousands of years of natural selection is

Seems like quite an over-generalization. I agree that a rational Bayesian approach would favor organic foods, because thousands of years of natural selection is probably a better indicator than laboratory tests of long-term low-dose toxicities and important but uncommon complex molecules in food supplies. But that doesn't mean that there aren't plenty of natural substances that are damaging to the human body even in moderation.