I agree. I'd also point out the oft-repeated point that "correlation does not imply causation".

My own hypothesis would be to conclude that "social support causes happiness" and that religion is not important at all except that religious communities have an historical association with social support networks.

Have you specifically LOOKED for a correlation? i.e. have you attended a church to see how its members behave? I won't say this is why I went to church, because it had more to do with a) liking the music, social atmosphere, etc and b) wanting to understand how someone could 'believe' things that seemed so irrational. (And now I think I understand that better.)