I agree with both of you that there are many atheists doing this sort of study that are (rightfully) annoyed by the Christian assumption that less religious (especially less Christian) people are less moral, and so want to say
"Oh surprise surprise, it's the opposite!"
And then of course too, being more religious doesn't equal less crime. I'd say firmly teaching particular ethics in a kid from a young age does that, which can be done with or without religion, ("God will punish you", versus How would that make you feel?", as examples) and it's better to do it without religion, since that encourages greater rationality and critical thinking later on.
Plus, if the kid later realizes the religion is bullshit, he doesn't also throw out all the decent or semi-decent things his parents taught him through the religion; don't steal for example.
You can see this with Christians who become "Satanists", aka devil worshippers aka reverse Christians. They just become "evil" as a form of rebellion. But I'm rambling a bit off course now.
Of course when we say "religious", well, there are many different religions. And religions change over time...modern Christianity is not Medieval Christianity or Ancient Roman Christianity.
I'd say for example, we have the rise since the 1960's of very liberal, tolerant Christians who are almost Unitarian Universalists, who have sort of an "All paths lead to God" feeling. These types are common in certain cities and towns, on both coasts of the continental US.
An individual can construe ancient religious texts to mean different things; for or against slavery, for or against expansionistic war, for or against freedom of speech.
But, if we look at what religion and then what "brand" of that religion is most popular in a given population, we can start to make some useful generalizations about the "Religious" and "Non-Religious" in that population.
True correlation isn't causation, and other things go on other than religious observance to determine levels of crime in a society, I said that.