I just read an interesting article that takes an empirical approach to assessing the practical validity of religion as it relates to socioeconomic success of a country. The results ended up saying more about economic policy than about religion, but it's still worth summarizing.
So what they did was track societal success via about four dozen measurable variables, everything from incarceration rates, homicides, overall health, divorces, to overall happiness. It's important to note that all countries used in the study are modern industrialized nations, as other factors would skew the data in other countries. And, across the board, there was a correlation with these measures of societal success and a lack of religion.
But the link, it turns out, wasn't exactly with religion. Rather, those that scored higher were those countries with progressive socialist economic policies. And, in turn, it was the countries' economic success that then drove down the religiosity (i.e. people turn to religion in stress/crisis/etc., so a more stable nation will see less of this and an increasing secularization). So the results were less about a lack of religion, and more about societal success leading to a less religious population.
You might think this is me posting this because it's a win for my personal beliefs. But it's not, because I'm actually quite libertarian in my economic opinions. And while I still think there is philosophical merit to more libertarian fiscal policies - which create personal freedom - most of my backing for the opinion is theoretical. Aside from case studies or smaller examples, the empirical justification for the practical benefit of it doesn't exist. So this is more problematic for me than it is affirming. Because what is a noble principle if it doesn't translate to practical benefit?
It also leaves us without sweeping empirical research to suggest that religion or non-religiosity is beneficial to society. The only evidence we DO have tends to suggest that those who are less religious tend to be more moral - detailed here - but this is on an individual level, not a societal one. In terms of large-scale empirical research, neither side has much to hang its hat on.
Interestingly, though, one argument that I've heard is that Christian/religious charities do more than non-religious charities. On an absolute basis, this is undoubtedly true because of the volume alone. Many more people are religious than not. On a per capita basis, I don't know what the data says. But the study also points out that there is markedly less charitable giving in more successful nations. This likely isn't from decreasing altruism, but because there is less need for it. When a population's basic needs are taken care of, and things like universal health coverage allow them more leverage in personal and professional realms, less economic stress, and less chance of going bankrupt, there is less need for charity in general.
The name of the article is "The Health of Nations: An Empirical Study on the Effects of Religion and Economic Policy" by Gregory Paul. I don't know of an online link to the study.