What I am wondering about is why it seems that atheists have complete caricatures of their previous theist beliefs.
Suppose there is diversity within a religion, on how much the sensible and silly beliefs are emphasized. If the likelihood of a person rejecting a religion is positively correlated with the religion recommending silly beliefs, then we should expect that the population of atheist converts should have a larger representation of people raised in homes where silly beliefs dominated than the population of theists. That is, standard evaporative cooling, except that the reasonable people who leave become atheists, and similarly reasonable people who are in a ‘warm’ religious setting can’t relate. (I don’t know if there is empirical support for this model or not.)
Suppose there is diversity within a religion, on how much the sensible and silly beliefs are emphasized. If the likelihood of a person rejecting a religion is positively correlated with the religion recommending silly beliefs, then we should expect that the population of atheist converts should have a larger representation of people raised in homes where silly beliefs dominated than the population of theists. That is, standard evaporative cooling, except that the reasonable people who leave become atheists, and similarly reasonable people who are in a ‘warm’ religious setting can’t relate. (I don’t know if there is empirical support for this model or not.)