> Many of those failures are capable of generating reasonable incomes for employees and founders for years, but those incomes are not proxies for investment returns.

That's part of my point. The current VC model causes failures of businesses that are viable going concerns, but not growing fast enough. They know they can make more overall by pushing a subset of those business to explode at the cost of killing most of the others. This also allows them to increase the velocity of their investments, further pressuring their startups to grow or fail fast. Their "cohort of pedigreed startups with plausible business plans and some degree of traction" should yield enough moderate successes to make investment profitable, as long as they are willing to play the long game.

Also, those incomes become proxies for investment returns once the company is stable enough to start paying distributions to its owners. Again, though, that ties up capital for the long term, which is anathema to the current VC crowd.