But the job of someone who wants to object is to offer a theory (no experimental results required initially) as to why pedophile sex offenders would be less intelligent than non-pedophile sex offenders but that difference would not show up among the non-criminal. With such a theory and the means to test it, the scientific process could proceed.

Dante already offered one offhand; and it isn't even his, but something that was originally posted at BC during one of his visits:

* Some crimes require greater intellect to commit than others.
* Rape requires particularly little intellect to commit.
* The weaker the victim, the less intellect it requires.
* Raping an adult requires more intellect than a 15yo than a 12yo than a 10yo than a 7yo than a 4yo...

This would not show up among pedophiles who never have contact and therefore are never incarcerated for contact; and most probably also not among pedophiles who have contact but only consensually (which should also be harder to catch and incarcerate than those who outright rape.) [And you know well what I mean here by the consensual/rape distinction even if you don't share my view.]

Certainly pro-contact activists have cited other species. It is a weak form of evidence because other species are far more different from humans than sex offenders are from ordinary people. But when arguing for instance that greater human male aggression is innate rather than learned, findings from a wide range of closely related species that all show the same thing are highly relevant. In that case there is a control group in each species, namely females.

Um?

Wrong example.

I understand your idea, but your example is still wrong.

A Homo genus specimen and a Pan genus specimen of the same gender are genetically closer to each other than to their respective opposite genders.

Yes, you miscegenate more in Homo+Homo or Pan+Pan heterosexuality than you would in Homo+Pan homosexuality.