What do you suppose we do?
For the long-term, here's the financial structure for the solution that I see: http://michaelochurch.wordpress.com/2013/03/26/gervais-macle... . Essentially, I think VC-istan and MegaCorps are both dead ends when it comes to genuine technical excellence (which weird people like us care about). I think it's time to have a serious conversation about financing a fleet of mid-growth K-strategist startups instead of these red-ocean, r-strategist, get-big-or-die marketing gambits.
Here's an insane thought that I had that just might work. Find a Midwestern city that has $25 million to blow on becoming a top technology hub and set up an Autonomy Fund. (I'll do the "grunt work" of screening for talent; if this idea has legs I'd drop everything to implement it.) 100 top-notch programmers, $125k per year (out of which their resource/AWS costs come, so no one's living high on the hog), and 2 years. First, these companies are designed to become profitable, not to exit, so compensation is profit-sharing, not the joke equity offered by VC-istan. The city that funds this takes a 37.5% profit-share of whatever they build (it's valuating their work at $333k per year, which is lower than a VC valuation, but the terms are better.) If it works, then add time to the schedule (i.e. more years of life and more startups) and possibly more engineers. This is like Y Combinator, but without the feeder-into-VC dynamic; it's to build real businesses that generating lasting value both to a geographical area and to technology itself.
I seriously think that Autonomy Funds are going to be big in the next 10 years-- VC-istan is essentially a shitty implementation of an Autonomy Fund, except with selection based on connections rather than technical ability-- but there are some obvious problems (moral hazard, principal-agent issues) that need to be solved.