I find this very interesting, and I hope that they get it peer reviewed somewhere so it's more "official" than arxiv. I don't normally view religion as a social activity, and this analysis seems to hinge on social factors (i.e. it seems to be counting edges instead of nodes). Religion doesn't seem to necessitate multiple people in the same way that language does, so I wonder whether their model might be fundamentally flawed. Looks like my pile of papers to read just got a couple pages higher.
The link to download the paper is here if anyone wants it.
I don't normally view religion as a social activity ... Religion doesn't seem to necessitate multiple people in the same way that language does
I believe there's some disagreement over whether religion is characterized as primarily a set of beliefs individuals hold, or primarily a sociological phenomenon. Clearly most religions have a mixture of both, so the disagreement is over which is more important or definitional. I think most sociologists tend to see it as more of a social phenomenon, with the role of shared beliefs being just one kind of social glue. Linguistically, we don't seem to call personal beliefs that one person holds, not shared by anyone else, a "religion"; religions seem to require some sort of community of members.
I fall more in the philosophical camp than the sociological camp with regards to religion, as I use the term to describe a subject of thought rather than a cultural institution, which is how your sociologists seem to be characterizing it.
I'm not sure that your linguistic argument holds in this case, because we're actually discussing the specifics of what "religion" means and so we can't appeal to standard usage. I, for instance, wouldn't call someone's personal beliefs on God a "religion" any more than I would call someone's personal beliefs on mental illness a "psychology" or their views on continental drift a "geography". In my mind, it just doesn't make sense for someone to possess a subject of thought.
If religion requires active participation in a community, then the existence of hermits who feel drawn to separate themselves from society to focus more on God seems to be a strong enough counterpoint that there's more going on. Like you say, there's disagreement about the term, but I'm just pointing out that this study only seems to make sense in a purely social way. People in these nine countries aren't going to stop thinking about God, death or any of the other usual subjects of religious materials, so I think calling for the extinction of religion is a bit premature.
I can't speak for the other countries, but private polling data in the United States reveals many more people who believe in a deity or who self-describe as "spiritual" than who are affiliated with an organized religious community by regular attendance at group meetings.
Christian scholars in the United States have noted that many self-described Christians do not have a set of doctrinal beliefs that accord with historic Christianity.
The link to download the paper is here if anyone wants it.
http://arxiv.org/abs/1012.1375