You'd be really hard pressed to find an evopsych study without basic errors in data collection or interpretation (usually both). Perhaps there are some, but most of it is, frankly, crap. Even if they know how to use statistics at the basic level (which is already rare), they mostly restrict their data sample to the WEIRD people, usually evopsych students.
While it is certain that evolution does not stop at the neck (why would it?), it is pretty certain that it changes there, and selects mostly for adaptability. Which makes sense. A physical body plan needs to be pretty much encoded in genes with little possibility of variation, a psychological reaction does not.
given how similar the bell curves in the physical and psychological realms are
Given how psychological studies are full of basic errors, and how few of them were successfully reproduced, we might be skeptical of whether there are these psycho bell curves real.
But even if they are, it still does not mean that a behavior they describe is universal (as most research in psychology is limited to the WEIRD), much less that it was selected for.
(OTOH a behavior might have been selected for which is a recent phenomenon, and thus limited to the WEIRD, but good luck finding one.)
This "nurture" explanation seems highly unlikely, and the social sciences have no testable explanation for it.
Now I'd like a single testable (not to mention tested) explanation in evopsych. Good luck with that.
Regarding your last statement:
Read the book Why Beautiful People Have More Daughters. It's full of tested predictions. Or read my post above again. I mention several, including the prediction that females should be the aggressors among species where males carry the babies. That's confirmed via observation.
Oh, and a Dobzhansky quote for you: