Uber could easily justify a $120 billion valuation by dominating American urban car transport. The big question is whether they'll be able to (a) dominate while (b) staying profitable.
Local network effects. Kalanick's strategic mistake was assuming network effects in New York create network effects in Hanoi. They don't, apart from cash.
But in New York, there are more drivers on Uber than other services. Which brings more riders. Which brings more drivers. Et cetera.
With their current business model, I don't think they have it in the bag. They don't have much that other ride shares don't.
But I think that if they had the infrastructure built out to support their current model for the whole country, they would be the only company capable of being the software side of autonomous vehicles. Or, at least, the best positioned company to win that contract.
I think that transportation itself is big enough to warrant such a big valuation. So I think their competitive advantage will be built over time and, right now, it's not anything sustainable by replacing taxis.
Of course, once someone wins the race, maybe we won't need anymore clones since the cost/transaction will reach a good equilibrium for everyone. Then they can build a different type of sustainable competitive advantage.
At least a few years ago: A willingness to break the law and then throw money at lobbyists and lawyers to change the policy.
But when they changed the local taxi laws/policies, they opened the doors to all ride sharing competitors as well, didn't they? If so that wouldn't necessarily count as a competitive advantage, more of a cost of entry to the market.