I certainly cannot prove that it's right, so how can anyone prove that it's wrong either, without setting up some type of psychological experiment? Therefore, I would classify the idea as a theoretical construct, useful for illustrating something about how people act, but not the type of thing you could easily say is right or wrong... For this reason, you cannot view anything in socionics they way you would in, say, the hard sciences. You can only see what ideas apply to you, use the ones that do, and discard the ones that don't... I'll admit you're not the only one here who is guilty of this. I would say that nearly everyone has made that mistake at some point in studying the theory, including myself...
Also, I apologize if I seem too hard on you. That's unfortunately the way this relation is... I would find it hard to communicate in any other way, and that's definitely a problem here...