FWIW, "evolutionary psychology" is a vapid autistic sort of speculation, devoid of empirical data or theoretical significance, but highly valued by those who would replace scholarship with armchair theorizing. We thought we'd outgrown this a century ago, but some ideas are just too bad to die.
Well, no, not necessarily. Suppose one accepts evolution as scientific fact; and anyone can observe that behavior seems to have at least some heritable aspects (e.g., golden retrievers tend to be sweet-tempered; pit bulls are dangerous). There would seem to be a plausible scientific basis for theorizing about "evolutionary psychology."
Of course, the problem (as you suggest) is that there is plenty of room for mental masturbation on the topic; and it's easy to let one's personal opinions (such as atheism) get in the way of theorizing. And of course, because "behavior" is an incredibly complex thing, it's beastly difficult to test any of those theories, and there's no particular penalty for not testing.
But Colson's complaint is similarly flawed -- he seems to insist that behavior would be required by this theory to be hard-wired to a particular cause/effect relationship ... something easily swatted aside by a Pinkerite, who would point out that evolution would most likely select against animals that only reacted one way to a variety of different stimuli.
Where Pinker's flaws result from his atheism, Colson's problem is that he cannot accept the idea that behaviors may be heritable traits, at least in part ... and both guys seem to want to treat something as incredibly complex as behavior, into nice neat packages to support their personal biases. It precludes a rational discussion.