Atheists may be smarter than the religious, but the religious are more likely to do the right thing when they aren’t being watched (see also Ron Bailey’s post here). On the other hand, the differences between the religious and nonreligious in these studies are unlikely to justify quite the levels of mistrust toward atheists found in U.S. polling data.
(An aside on methodology: The commenters on the last link seem to discount these studies automatically on the assumption that data can always be manipulated. [I know I’m giving a few anonymous blog commenters more than their due, but all the same…] If that were true, then we should just shut down all empirical research in the natural and social sciences. That data can be manipulated does not mean that we are justified in assuming that it is. From a Bayesian perspective on social science theories, we should at the very least take these data into account in updating our beliefs, even if we do not regard the findings as conclusive. Nonspecialists too often take social science research as all-or-nothing: either a study is worthless, or it Proves Everything.)
PILEUS 
As a Christian, I would submit to you that the religious never feel that they aren’t being watched.
Yes, that’s exactly what the researchers involved in these studies hypothesize.
Only atheists are comfortable manipulating study data.
Cause they don’t think anyone’s watching.
You neglect the long history of document forgeries in the name of religion: Donation of Constantine, Pseudo-Isidorean Decretals etc. Or were you joking?
Absolutely, although it could be true.
Benjamin Franklin figured this out almost three hundred years ago. All of his atheist friends took advantage of him, and he ended up revising his belief system on the pragmatic basis that some respect for the ethical component of revelation was necessary for felicity in this life.