So many religious facts have a very long half-life for their relevance. Say you learn about how the four Gospels differ — that’s still relevant for understanding Christian divisions or Christian theology today. Reading about the Reformation? The chance of that still being relevant is much higher than if you were reading about purely secular divisions in internal German or Swiss politics in those same centuries.
Jews, Buddhists, Hindus, Sikhs, or Muslims? Facts from many centuries ago still might matter. And the odds are that people a few centuries from now still ought to read about the origins of Mormonism.
In few other areas do past facts stand such a high chance of remaining relevant for so long.
As an empirical matter, “rationalists” tend not to read so much about religion, but that is precisely the unreasonable thing to do.
If you’d like to see a potential counter, here is some poll evidence that many people don’t care so much about the divisions of the Reformation any more. It still matters a great deal whether you are in Catholic, Protestant, Orthodox, or “mixed” Europe.
I am indebted to a conversation with Bryan Caplan for the main point, though he is not liable for my formulation.