Basic economics states that when many firms enter a market, the market becomes more competitive and the expected returns to those firms go down, which acts a disincentive to others entering the market in the future. The effect isn't immediate (entrepreneurs have imperfect information about the real returns in a market), so that's why you get boom/bust cycles.

Slightly less basic economics states that when a market becomes commoditized, its complements should raise in expected returns. So we ought to be seeing higher returns for web frameworks, scripting languages, open-source databases, scalability services, web hosting, startup boot-camps, and services that help people make sense of the Web 2.0 clutter.

I think we're seeing this, but the problem is that other folks realize this too, and so we're seeing people pile into lifestreaming apps (eg. FriendFeed) and YC clones. I'm not really sure what industry has the best returns now - if I was, I'd be rich. I suspect it's not Web2.0, but there's something within tech that will be very, very high returns over the next 10 years, but it's not an obvious industry at this point.