The general consensus in the field is that 4x is an overestimate for the reasons you describe, but the "ground truth" would still be increased prevalence in men

The general consensus in the field is that 4x is an overestimate for the reasons you describe, but the "ground truth" would still be increased prevalence in men. Interestingly there are even some mouse models of Autism where the male behavior is what one might expect from an ASD model, but the females don't really behave differently from controls. There are of course also some models where you can get erratic female behavior. Ultimately I think there's just a lot we don't understand yet.