The human experience doesn't change; not until the fundamental biological constraints of being human change.
This is silly. Human experience is dependent on much more than just fundamental biological constraints. Society / culture are huge factors in our behavior and development.
Surely culture changes. Do you think that 18th century Japanese culture has any fundamental differences compared to now? How about compared to 2000 years ago? How about compared to 20,000 years ago?
Do you think religion was a factor for the pilgrims who first came to America? How about 1000 years before that? How about 20,000 years before that?
If we don't need religion now, then we never needed it. If its a dangerous influence now, it always has been.
Culture, Crash. It's all about culture.
Here's another good example of how cultural changes dictate how things may be cohesive or may be divisive: nationalism. Do you think nationalism brought a country together after 9/11? Do you think anything has changed since then? I would suggest that nationalism was a big bonding factor around 9/11, but now is turning into a divisive force. Really fast change.
Culture changes. Needs change. Knowledge changes. The role of religion changes.
Religion serves to divide the world into believers and unbelievers. In that sense, I would describe it as an influence against cohesion, a divisive influence.
Not if everybody's a believer. Not if believers are stronger from their faith and are able to co-exist with non-believers.
Maybe how religion works is dependent on population size of a culture. Paper didn't investigate it, and you're dismissing the possibility without considering it. Maybe religion works really well for tribal-sized cultures. A good place to examine might be in some African nations then. What are the cultural mechanisms used to keep a tribe together as a cohesive unit?
Even if religion provides ceremonies; music; leaders--then as far as these items are providing cohesion, religion is too. People aren't as autonomous as you're wanting to make them out to be.
Ben writes:
Do we really think that religious belief doesn't correlate with other important behavioral factors that might invalidate their findings?
It's a chicken-egg situation only if the correlating factor is causally connected to religious belief. If it's only incidentally related, then it means religion has nothing to do with the correlation.
For example, let's say that test scores (crimes) correlates with skin color (religion). Let's also say that skin color (religion) also correlates with income (unknown factor X). According to you, it's a "chicken-egg" situation between test scores and skin color. What I'm saying is there's a 3rd external cause--income--that relates the skin color (religion) and test scores (crime). Thus, skin color (religion) has nothing to do with test scores (crime); it's an incidental correlation. It's possible to have one without the other.
There's so many things religion could correlate with; simple ones that have no causal connection but rather incidental, such as geographical location. You can't make the conclusions you want to unless you rigorously explore these scenarios.