I think the paradox is that having a very good CEO hypothetically is worth a premium for businesses with massive revenues: Where even the square root of a profi

I think the paradox is that having a very good CEO hypothetically is worth a premium for businesses with massive revenues: Where even the square root of a profit delta relative to a 1% less effective CEO is huge, it makes sense that it's worth a nonlinear increase in pay to get someone who's 1% better. But simultaneously people are very unlikely to correctly identify which person of a group of competent candidates actually is the best. It's just that marketing and confirmation bias are good enough to cover it up.