You have suggested that silicon valley is successful, in part, because workers accept a notion of "long-term compensation" [1]. long-term compensation, perhaps better called maybe-someday compensation, requires sv software engineers to work for relatively less (factoring in cost-of-living) than many other skilled professionals, with hopes that they'll be rewarded with a big windfall in the future.
with the rising cost of living in sf/sv, young engineers are expected to shoulder much the risk of startups while putting their long-term financial stability on the line. all in hopes of maybe-someday being compensated. and let's face it, the lion's share of these rewards will not be passed to these workers who put all their eggs in one basket, but to financially-secure vc's who have hedged their risk across many companies. if this is how sv works, do you truly believe this is a strong foundation for the industry? obviously it's good for founders and vcs, but i'm considering the perspective of engineers.
we seem to be enjoying go-go years of investment and profits, everyone is starry eyed, and the system is working. what happens when the present-day strain on workers exceeds the allure of maybe-someday compensation? also, why shouldn't talented workers move to an industry that provides guaranteed compensation for hard work in real time?
of course, start-ups have a much different risk profile than banks, consulting agencies, large corporations, etc. and as a result, the long-term compensation model may be a necessity. but in the end, it sounds like sv works because there's a large supply of people willing to gamble on their future financial stability.
[1] http://blog.samaltman.com/why-silicon-valley-works