Have Reason and Religion Evolved by Natural Processes?

I heard a debate between Alvin Plantinga and Daniel Dennett on the way home from work the other day. According to the debate organizers at issue was the question, “Are Science and Religion Compatible?” But as the debate actually unfolded it seemed to center specifically on Plantinga's "Evolutionary Argument Against Naturalism."  

Now it should be noted up front that I am personally quite skeptical of what I call grand-scale evolutionary theory: descent with modification of the entire range of presently observable biological diversity, by strictly natural mechanisms such as selection, and from a single common ancestral stock. Call me personally incredulous. But I did think it fascinating that even on the premise that grand-scale evolution is a “fact of science” that no rational person would ever doubt, Plantinga has nonetheless perceived a self-defeating weakness in the popular philosophical conjunction of naturalism and evolution – namely that the probability of our cognitive faculties being reliable, given naturalism and evolution, would be low (where “low” presumably means either less than .5 or less than the probability of reliable cognition on some other view).  

The way I heard Plantinga he argued basically this: The criterion for evolutionary success is survival, not truth. So there's no reason to think that a highly evolved organism would be adept at tracking truth, rather than simply surviving by whatever means necessary. The example he gave was of a frog whose instincts have evolved so that it shoots out its tongue with pinpoint accuracy to ensnare a passing fly. Whether the frog has any “beliefs,” or whether they are true, appears to be irrelevant to the frog’s short-term survivability, hence potential reproductive success.  

Dennett, again to my hearing, argued to the contrary: You silly theists don’t understand the role of cognition in determining evolutionary fitness. Evolution works as organisms adapt to their environments, and this means, where applicable, gaining the cognitive skill necessary for survival. Naturally, improved cognitive skill in principle translates to an improved grasp on the truth in practice. I can't recall if Plantinga mentioned this or not, but Dennett's counter, based on the proposition that humankind now possesses highly rational, thoroughly reliable cognitive faculties, leads Dennett where he would rather not go. That is, given that evolution has produced reliable cognitive faculties, and given that humankind is highly religious, the probability of religion’s being true is very high.  

So a recap might go something like this:

THEIST: Evolution cares nothing about truth. Hence an evolved cognition may function properly and still believe what is inherently false.

NATURALIST: Yes, but cognition only evolves if it improves. More improved cognition by definition tracks truth more reliably than less improved cognition.

THEIST: Okay, but evolution has led a substantial majority of humans to embrace religion in practice. If we agree that human cognition has improved, and improved cognition implies a better handle on the truth, then religious faith has a better handle on the truth than naturalism.

At this point I suppose the atheist could still protest, albeit feebly: But religion is in decline.

That all leads me to the question: Given evolutionary naturalism, just how far along are we on the grand scale that is evolutionary history? If we are still evolving, as evolution would seem to suggest, tomorrow our cognitive faculties will very likely be more evolved than today, in which case the beliefs we hold to be rational or veridical today we may consider irrational or absurd tomorrow. Of course evolution has no ideological agenda. So beliefs theoretically liable to become evolutionarily obsolete include not only religious beliefs, but beliefs in the validity of logical and mathematical axioms, and belief in naturalistic evolution itself.  On the other hand, if human cognition reliably tracks the truth and has reached its apex, it seems to follow that some form of religious belief is much more likely to be true than a belief in naturalistic evolution.