"At the moment, there is no one within big companies who gets in trouble when they buy a startup for $200 million that they could have bought earlier for $20 million. There should start to be someone who gets in trouble for that." - Paul Graham
I think its well understood that there exists a tremendous amount of inefficiency. That large companies should spend some energy identifying what they need so it costs them 20 million instead of 200. I'm know some large companies already try to do this. They will get better at it over time and others will emulate their models. As a result, the investment and startup model/process/vc-lifecycle will shift accordingly. I feel PG thinks that ycombinator is on the leading edge of being able to take advantage of this shift as...he probably is.
My attitude for my startup has always been that if I sell early, I get 20 million and get to keep all of it. If I take on lots of VC, I have to sell for 200 million and I still net out at 20. Most stertup founders are happy to take the early 20 and get on with their next big idea. I do think PG is correct that the model is slowly shifting in this direction. And its the right direction for innovation to continue.