Stanovich draws an interesting distinction between intelligence and rationality, where intelligence, as measured by IQ test, is so to say the strength of an individual's analytical abilities, whereas rationality is this individual's tendency to use these analytical abilities (as opposed to fast and unreliable Systems 1 processes); i.e., his or her tendency to "overcome his or her biases". According to Stanovich, there are large individual differences not only regarding IQ but also regarding rationality, and he is now in the process of constructing a test measuring people's rationality quotient, RQ. Now my question is this: in which areas do you think that a higher RQ people have comparative advantages, and in which areas do you think that high IQ people have comparative advantages? My hunch is that IQ pays off better in precise fields like mathematics, physics and computer science, where the problems are often so hard that most people can't solve them even if they overcome their biases and use their System 2, whereas high RQ pays off better in more ill-structured fields like qualitative sociology, where any individual line of reasoning usually is fairly simple, and therefore does not require a very high IQ, but where it is easy to fall prey to (politically) motivated reasoning, confirmation bias, and all sorts of other biases.

Hence in order to arrive at true theories, it seems to me that you need a high RQ in the social sciences. On the other hand, in order to sell your theories, RQ is not necessarily always helpful: on the contrary, a fair dose of overconfidence bias can be useful here. Many bigshot social scientists during the last century or so were anything but rational (Foucault and Freud are two of many examples), but were able to convince other (equally biased people) that they were.

As fields become more exact (as for instance psychology gradually have become), you gradually need a higher and higher IQ to compete: rationality is no longer enough. My guess is that as more and more fields grow more exact, moderate IQ people will be of less and less use in the academia.