I left a really long comment that probably should have been posted here

on Mark Shea's blog about this story.
Here's what I said, if you don't want to click:
It's sad, though, because lots of mis-educated people don't understand the problems and think that "Science" has really concluded something here. I had a nice atheist I was corresponding with send this to me as though it represented something of importance. Of course, had the scientist been the sort of theist to conclude that the data meant that God was punishing the U.S. for its sins, no one would have given this "research" the time of day.
Just to spell it out - scientists rightly look on meta-studies, or studies that merely aggregate the data from other studies, as significantly less revealing than individual studies with large subject sets. One reason for this is because in a meta-study, such as this one, the researcher has all the results at the same time he designs the methodology of the study, which makes it difficult to be objective. For example - why does this study compare abortion, suicide, and murder rates? I'll bet my left leg it is because those are the areas in which the U.S. compares poorly with other nations, and a researcher who wants to show that the U.S. is hypocritical can selectively choose these criteria. I'm sure there are other criteria on which the U.S. looks better (or even other sets of studies, or subsets of data within the studies used - what years are compared, for example) but these are simply not included, since that is not the point the researcher wants to make.
And of course there are thousands of reasonable explanations for divergences in these rates that don't include religion. And of course he is stepping wildly out of reasonable interpretation of the data in concluding philosophical claims. But Mark already made that point.
I'm sure there are decently conducted meta-studies. But these "studies" are cheap and easy to fudge in these ways, hence their popularity.