Jonathan Haidt has a reply to the replies to his essay "Moral Psychology and the Misunderstanding of Religion", and he pretty much misses the point of the criticism about the truth value vs. the benefits of religious belief. Here's what he says:

But what if we drop the methodological individualist criteria and just ask about the degree to which religion makes people divert their time, money, and attention away from themselves? The charitable imbalance between atheists and believers now becomes enormous (according to Brooks), and the analogy I and others have made between religious communities and beehives becomes more useful. All that time and money given to one's own church is like the "altruism" of bees who toil to build their common hive; all that time and money given to build churches in faraway lands is like the efforts of bees to found new colonies. Religions, generally speaking, work to suppress our inner chimp and bring out our inner bee. But methodological individualists, who deny group-level selection and shun group-level analyses, find it hard to believe that people could be happier or more generous when they live in bee-like ways than when they live on their own, outside of any hive.

Two points: One, Haidt misses the point that PZ Meyers makes, that even conceding that religion makes you happier and healthier, that's not a good justification for believing it. Here's Meyers:

By my side right now, I have a small plush animal. If it were conclusively shown that beliefs in a god or religion were definitely beneficial in and of themselves, that humans needed this little kernel of worship in order to thrive a little better, and I said that my toy octopus was a god, lord and savior of us all, and if only you believed in him, you would gain an empirically demonstrable extra year of life and a quantifiable increase in your happiness, what would you do? Would you abandon one little piece of rationality and bow down before the toy? Would you even be capable of that level of credulity?

Sam Harris makes the point this way: Believing that you have won 12 Olympic gold medals in sprinting would probably make you feel happy and give you more self-confidence. So would you be justified in believing you are the fastest person in the world, regardless of whether or not it was true, if the benefits were tangible?

Second, Haidt advocates a system which "suppresses our inner chimp and brings out our inner bee", but seems to be implying that religion is the only way to go about that. You don't necessarily need to believe in supernaturalism to behave communally and cooperatively.

As I've said before, his criticism that the New Atheists seem to be engaged predominantly in the tearing down of religious beliefs without offering up substantive replacements for morality and meaning is legitimate. But his continued defense and reverent politeness toward belief systems he acknowledges are unfounded and often harmful is unjustified...unless he professes to value happiness over truth, a very unscientific, but very human, thing to do.