Even if winestock compiled a complete list of 195 countries, their religiosity, and their fertility rates, and showed a correlation between them (which it seems is unlikely -- see for example steadicat's post, or my other post), it would still be wrong to conclude, as winestock did, that

> Lack of religion is either non-adaptive from a fitness perspective, or it is strongly tied to traits which are themselves non-adaptive.

There is at least one other possibility, which is that there is no link between religion and fertility, even transitively.