All the claim about the highest IQs being theist tells me is that the author doesn't really understand what the IQ measure means - and, at best, it tells you that even some really clever people can be theist
When discussing group differences in IQ, we should, by definition, be looking at averages - that's what IQ is about. Claims about the highest IQs in the world are pretty meaningless unless you've randomly sampled the same (large) number of each group, correcting for or equalising all the relevant influences (poverty, education, culture, etc), which is practically impossible. This is why IQ is intended for, and most meaningful when, measuring relative differences within the population for which it has been calibrated.
The interesting question to me is why the surveys suggest that atheists have higher IQs - I suspect it's a question of sampling error and/or self-selection.
For example, assuming both atheist and theist groups start with average IQ and a normal distribution; if atheists are a minority and an equal percentage of the atheist and theist groups switch 'allegiances' in a given time, a greater percentage of atheists will then be ex-theists than theists will be ex-atheists (because of the difference in group sizes). If those that switch either way are more likely to have a higher than average IQ, that will push up the average IQ of the atheist group more than that of the theist group. So even if all other considerations are equal (and they won't be), if those that switch have slightly higher than average IQ, group size difference alone will cause the smaller group to increase IQ compared to the larger group.
But there are plenty of other reasons why such a difference might be found.