I hope this doesn't get entirely buried at this point, but I think you bring up something interesting with driving marginal costs down.

Ostensibly both Uber and Lyft are doing the same thing, but with the item that will bring the marginal cost down being self-driving cars. They've both bet big in order to capture the market, because it's likely that whoever owns the market before that transition will own it afterwards as well.

It does seem like there was probably more optimism that the technology would materialize sooner than it really has, but assuming it materializes some time in the near future it's likely that whatever company manages to hold onto the market until then will reap massive rewards.

It's clearly a risky bet, but at least from my view doesn't seem like a pump and dump type of scheme.