Across 5 funding rounds, Crunchbase lists Loopt as having raised $39 million and then was acquired (acqui-hired?) for $43 million. He didn't create any multiples of value for his investors. Loopt wasn't a breakout hit like so many other YC startups have been. It was certainly one of the first interesting location-based apps in the App Store, but soon was surrounded by other location-based apps and never really appeared to surface and gain traction.
Obviously Sam runs YC now and has dramatically improved it, but in the lens of being an entrepreneur, isn't he still essentially unproven, and not a success story in the startup world?
But, of course: if the model they're building with YC works, this is the outcome you'd expect. YC wouldn't be very interesting if its outcomes were capped to Viaweb's and Loopt's.
Paul Graham was a borderline founder of Reddit, as well.
Not saying I agree, BTW - I think that binary startup success has a large skill factor, but the magnitude of that success is largely luck, and that good luck can mask bad practices to a large extent. So in my view, people with relatively minor startup successes like Sam Altman, Paul Buchheit, or Michael Siebel might actually be better coaches than knock-it-out-of-the-park successes like Larry Page, Mark Cuban, or Notch. But don't misrepresent the argument: operational experience running an existing business is very different from experience founding one.