Collier is merely smart. He's very, very smart indeed. But in the domain of music, expressive originality is far more important.
And it can be created by people who aren't technically music-smart at all. It's a different skill to the kind of grammatical/technical smartness that someone like Collier has.
That's not a distraction at all. It's exactly the point - creativity is orthogonal to smartness, and it's poorly understood and even more poorly supported.
One of the interesting thing that happened in English pop (until it stopped being possible a decade or two ago) was that successful pop artists were more likely to have been through art school than music school.
The exemplar is Eno - who studied with Roy Ascott, who is probably one of the most unknown influential pioneers of computer art.
Eno has taste and a willingness to experiment across multiple domains. The rest almost falls into place.
IMO the combination of openness to experimentation and instinctual feel for rightness/wrongness to guide that experimentation is the foundation of useful creativity.
Art school - more than anywhere else - gives people permission to experiment. Taste can be partially taught, but you need an instinctive feel for it, and that's probably innate.
Someone like Collier doesn't score highly on either experimentation or taste. He scores very highly on musical competence and technical skill. But both are a kind of conformity - which is the opposite of real originality.
I assume where PG is going with this is the suggestion that some people can originate incredibly successful business ideas, and most people can't.
Which might be nice if it were true. But in a startup sense the opposite is more likely. You need a baseline of conformity to be in business at all. Truly creative types don't find the business world open enough to be interesting.
Success in business relies on having ideas that are original enough to be different, but not so original they're incomprehensibly challenging and difficult.
Market fit mostly happens near the middle of the bell curve - for whatever metric you're measuring - and that's not where the most creative people like to live.