With M&A as the only exit available, the pay out for any employee other than perhaps the first-five isn't going to be a "home run". Thus there would be nothing for them to justify giving up a higher salary, stability and working longer hours: their options won't be worth much.
Trying to grow the company to post-SOX IPO size would also mean a much greater gamble and much greater time investment. Now the options wouldn't account for anything for possibly as long as ten years (which isn't much shorter than climbing the ladder at a big company).
This means that smart hacker's most economically rational choice is either working for a big company (which issues RSUs or options for publicly tradeable stock) or co-founding or being employee# 1-5 in a less ambitious venture (which usually isn't going to provide very much of a technical challenge).
Not all is bleak, however: it's unlikely that SOX will go (but perhaps I am a pessimist when it comes to govt. getting smaller) but we could see mid-size private companies providing revenue sharing/bonuses (in lieu of options), smarter M&A strategies (e.g. acquirers letting the companies they bought maintain their culture and grow, while offering employees additional upside or even spinning the ventures off for an IPO - but with big co's resources, most notable example of this is EMC + VMWARE acquisition and IPO) and finally markets where smaller players could still have interesting technologies to play with (vs. general database driven websites, which don't really involve much "hard core" tech).
It could be happening via investors, though. It could be that e.g. Google decided to go for the big time partly because there was so much funding available in 1998, which in turn was true because investors were hoping for IPOs.
You haven't seen them say it directly, but how many applications do you get that are for fairly ambitious and big projects? What percentage are front-end heavy web-apps/aggregators? What percentage are fairly ambitious and require a serious research, development and have a huge potential for growth? That's a question you may be able to answer better than anyone else (although I'd imagine most YC applicants are a self-selected crowd and those hoping for an exit other than M&A and to hire a great deal of talent may not even apply to YC).
I also didn't think about the investment angle at all. I wonder how much investors are pushing for companies to sell quickly vs. grow.
I'd say many if not most of the startups YC funds have the potential to go public, if the founders were sufficiently driven. (In saying that I'm relying on the fact that practically no companies that go public look like they will at the start. Apple started out as a company that was going to sell plans for computers.)
Wow that's something I haven't entirely considered, despite knowing it to be true. Somewhat adds to my argument, however, without the chilling effects of regulation we may have seen a more interesting and diverse market -- but now I am in purely speculative realm.