If it's worth saying, but not worth its own post (even in Discussion), then it goes here.
The "control" argument predicts more specific things than the "rebellion" argument, and so is a more useful hypothesis.
Actually a rebellion argument also predicts something. It would predict that atheists also rebel against other social norms.
Because the competing hypothesis ("atheists are willing to state a true thing even when most of society disagrees") also predicts some degree of general rebelliousness, I think the prediction is more about pointless and self-destructive behaviors.
And if atheists are just allowed to be tricked by the devil, then I don't know how that pans out into other behaviors.