I'm not sure if I would buy this conclusion without more data - it's quite possible that higher-IQ types are in qualitatively different roles than lower-IQ type

I'm not sure if I would buy this conclusion without more data - it's quite possible that higher-IQ types are in qualitatively different roles than lower-IQ types in the same study. This would be very difficult to control for because if you do control such that you're comparing two individuals who are in similar roles but different IQ, you're then implicitly selecting for people who have other compensating characteristics.