Like most things in real life, the answer boils down to a set of intertwined non-linear feedback loops that you can probably identify and analyze (even quantitatively) using systems analysis (http://en.wikipedia.org/wiki/Systems_analysis) or a similar technique. I suspect pg in his essay has nailed down some of (and probably the major) loops involved, i.e. the ones around rich people and nerds.
The basic loops he identifies seem to be nerds attract other nerds, rich people attract other rich people, and nerds attract rich people as long as they are starting startups that need investment and get rich people richer (please correct me if I am wrong). These are all positive feedback loops, so they keep growing until they hit some negative ones. Some negative ones: saturation of nerds or rich people (not yet), saturation of space (getting there), high living costs (definitely close), nerds not starting startups because they work for Google (hard to tell but I doubt it), nerds not needing investment (increasingly so for Web apps), and other viable places to move (true in some areas like Biotech).
This is the kind of tangible analysis that I was looking for. I must giveaway that I'm my last degree was in systems engineering, so I completely understand this. Thanks!
The basic loops he identifies seem to be nerds attract other nerds, rich people attract other rich people, and nerds attract rich people as long as they are starting startups that need investment and get rich people richer (please correct me if I am wrong). These are all positive feedback loops, so they keep growing until they hit some negative ones. Some negative ones: saturation of nerds or rich people (not yet), saturation of space (getting there), high living costs (definitely close), nerds not starting startups because they work for Google (hard to tell but I doubt it), nerds not needing investment (increasingly so for Web apps), and other viable places to move (true in some areas like Biotech).