So are you saying that that's not actually the case, that smart people aren't disproportionately unpopular? Interesting if true. Definitely needs some kind of justification, since it flies in the face of most people's intuition.
I'm not sure what the answer is, but I'm leaning more towards the position that intelligence is negatively correlated with popularity. Smart people aren't necessarily outcasts, but IMX you'll tend to see fewer of them at the top of the popularity curve and more at the bottom than you'd expect from a random distribution. And I imagine that there are a lot of reasons for why this is.
PG's argument in particular is closer to the truth than it seems at face value, I think. He says that smart people's problem is that they don't spend enough time on socializing; I'm not sure if this is really the prime cause of smart people's ostensible unpopularity, but it applies to my experience in middle and high school. My parents were rich enough to send me to good private schools that valued intellectualism (which probably more closely parallel schools outside the US) so I was never really unpopular, but I was definitely less popular than I wanted to be. In retrospect it's pretty clear that I could've been way more popular if I'd doubled or tripled the amount of time I spent on it; instead, I spent my free time programming and doing math.
The only place that I disagree is with PG's rationalization of why nerds don't socialize more. In my case, it wasn't that I wanted to build great things (although to some extent I did). I'm pretty sure it was just that solving math problems was a much more reliable and straightforward endorphin trigger than socializing.
Anyway, all this is to say that a) I disagree with you and b) you should be less pithy and more explanatory.