That's not obvious to me at all... my rough (highly uninformed) model of how cult initiations work is that you get sucked in to a situation (e.g. Scientology "stress testing" or similar) where you're forced to either not conform or join the cult. (Also fits my (very limited) observation of fraternity initiations.)
Also keep in mind that people who join cults are not a random sample of the population. It is generally believed that people who join cults are usually in some way social outcasts to begin with. Cults offer them the possibility of being part of a close-knit community. Traditional religions also offer that, but to a lesser extent.
Just because a story is flattering to your group doesn't mean it's false. The research I've seen on the usefulness of high IQ for achieving desirable life outcomes has been pretty positive.
References?
AFAIK, it is well established that IQ positively correlates with many performance metrics for levels of IQ around and below the average, while the correlation for above average IQ is more dubious. In fact, IQ tests were originally designed to detect underperforming individuals.
The correlation between an high IQ and being a "nerd" is also debatable, while, at least in some circles, nerds may have an high IQ, it doesn't follow that many high-IQ people are nerds. Also note that the whole concept of being a "nerd" migh be largely the effect of conformity biases.
References?
These weren't what I had in mind originally, but they look reasonably good:
Here is a new post at EconLog in which Bryan Caplan discusses how signalling contributes to the status quo bias.