I think if the VCs look at what they actually do vs what they say they do, they'll see they're more like the movie industry. Funding a lot of ideas, without any clue which ones will be big hits. That suggests that the YCombinator approach is more viable. And I gotta say I
argued for that in the 90s, and was told (again) that I was hallucinating. I should have taken that as a compliment. :-)