But even in the '90s tech boom in Texas, many were using equity in tech companies as leverage to buy and develop land. This in a place with more affordable housing and no water bounding it.
What passes for capital in SF is also not necessarily easily understood as capital without extending the concept quite a bit. It's closer to speculation in early phase equities, which is about the ... third derivative of actual capital. Unlike a building full of machines that add great creative value but are hard to move, this brand of capital is many orders of magnitude more ephemeral.
Capital also is an input, not an output, so I'm not sure what distinction you are trying to make.
What he's claiming is that in "normal" businesses those who supply the inputs (land, labor, and capital) all get a decent share of the profits, but that in the digital economy most of the profits go to those who supply capital.
The problem I see with his claim is that he doesn't actually make a case that this is inherently bad.
He uses Airbnb and Uber as examples to illustrate how it can be bad. I agree both of those companies are bad...but they are bad because their business model is essentially based on turning many of the costs of the hotel and taxi business, respectively, into negative externalities, not because of how their profits are distributed among the suppliers of land, labor, and capital.
[0] Entrepreneur is a concept that AFAIK economics does not have a good way to deal with. It seems to be a special class of labor because entrepreneur make there money by knowing exactly what kind of labor would be most useful. But it also seems to be a special class of capital because they make all of there money on equities (by selling shares in their own companies that they bought for pennies at founding). If anyone has a coherent theory of economics that accounts of entrepreneurs, I'd love to read about it.