The problem with this philosophy is that it would have caused investors to miss very successful investments. Since the number of big wins is limited, there is a very real risk to professional investors to NOT being in wins.

This approach also basically rules out entire classes of investments (e.g. biotech).

It may be apocryphal, but Google (market cap: $567B) is rumored to have taken its first investment before it had incorporated, and well before it started making money. I'm sure there Google is not alone in this respect.

That's sort of antithetical to the purpose of venture capital, which I understand to be moonshot research projects that create something new, something that might take years of work before it sees revenue.

This misses the point.

The whole point of investing is that you are doing so based on potential. Regardless of what the current value is. In fact, the lower the current value is the better!

The fact that the pendulum has swung too far doesn't mean the approach is spurious.

Founders can invest their own money first, though. The only caveat is like some above who mention investments in research-level projects (not software startups). Depends on the company, but in the context of Zenefits they clearly blew the extra cash on nonsense.