You're right -- there's no disputing that.
However, the manipulations of these studies aren't looking at behavior. They're looking to score political points by mixing behavior (35% of molestations) and self-identity (2% is homosexual).
More people behave homosexually than identify as "homosexual".
Or, to continue your example, there are more people committing jailable offences than there are people who are "inmates".
Define "homosexual" then. Is it not one who practices homosexuality? At minimum, it is one who desires to practice homosexuality. Wouldn't all who practice it first desire it?
On the other analogy: If you commit a crime, just because you are not in jail doesn't make you less of a criminal. Similarly, claiming to be innocent doesn't make that true either. (And being accused doesn't make you guilty if you truly didn't do it.)
My point is: Truth is an observable absolute. Behavior is an observable factor. It is what it is. To find absolute truth you cannot deny uncomfortable, observable factors.
Feelings, on the other hand, are not as easy to know or observe. What one feels is not always related to truth. Emotion is pliable by experience as well as expectation of future outcome. Science is totally reliant on accurate observation of relevant matter. Any adjustment of premise will most certainly produce skewed results.
Behavioral science is a subjective science because it mixes the observable (deeds) with the subjective (why?). I am not saying it does not have value, but it does not have a lot of absolutes. Anatomy and biology, on the other hand, are precise and observable. When human behavior does not match human biology, the so-called science of behavior will never produce accurate results (being a subjective study) if it does not acknowledge the clear and observable truths of the other human sciences.
To claim that a person desires homosexual relations with a child is not homosexual in "orientation" is a complete contradiction and manipulation of clear meanings of words. It forces the "scientist" to draw conclusions by allowing feelings to trump facts and observable deeds. That's nonsense.
If a behavioral study manipulates the premises in such a manner, it can never draw an honest conclusion.