I do think that past successful collaboration can lead to a bias though. As a programmer I've solved problems for a variety of disparate industries and fields. I know just enough to know I'm vastly ignorant of those fields save for a little part that I helped solve problems for. But I've seen colleagues go the opposite route.

I have just observed that the threshold is low enough that many middle-class engineers achieve it (I've always lived in low cost of living areas where engineers get to live pretty comfortable lives relative to their surroundings. But that's just my one data point). Their folly tends to be more banal, like ordering Ivermectin from Mexican pharmacies or endlessly talking about nuclear power in a near utopian fashion. Admittedly, pretty harmless stuff compared to the article.