Methodology issue of Evolutionary Psychology

From my last discussion of the theoretical issues of evolution as a hypothesis to be tested, it follows that any disciple derivative from it would suffer from the same issue. This is an example I remembered from the heyday of my Psychology class in University: It is about the idea of biological preparedness for learning in Behaviorism, the central tenet is that Phobic is a sort of evolutionary ‘leftover’. Therefore it is easy to learned and difficult to unlearn. Which to me it seems to implied part of causation chain is in the genes.
How to prove this assertion? That is the part make this example so memorable. The method is to ask ‘experts’ to judge how each phobic is related to the level of danger posed to human five thousand years ago. Thus, follow this strange logic, no wonder human is so easily to acquire fear of height, fear of snake, fear of spider! Because some of our ancestor has suffer from them therefore develop an inborn preparedness to learn to fear them. It is adaptive(during that time) to born with tendency for the fear of them since they will increase your chance of survival.
Does it sound like circular logic to you? My feeling at the time is: That is certainly a circular logic. How can anyone setup an experiment to test this ‘hypothesis’? Can anyone setup a null hypothesis to test against it? How can anyone to have an experiment to falsify this hypothesis? If this hypothesis is right, you certainly can’t find another human that doesn’t have this inborn tendency.(If this hypothesis is wrong, how can you distinguish it from other?) What can five or tend ‘expert’ judge give us useful information on the past? At least, we require anthropologist to test if the native who live in a native environment who are also born with these tendencies. Otherwise, how can we know for sure? Evolution is a retrospective science, thus by nature, it is difficult to assure of its internal validity, so do Evolutionary Psychology.