http://www.nakedcapitalism.com/2016/11/can-uber-ever-deliver...
However, in a given city, I would expect there to be economies of scale as you optimize your cars better and decrease the amount of time they have no passengers. It's most clear to me in the case of UberPOOL where the more passengers you have in total, the more likely two of them will be going in the same direction. We may not have reached a cheap enough price point yet to hit these economies.
Depending on how much you trust anything Uber says, they say they are profitable in SF, and some other US metros, which the article you linked ignores.
I think people are right to be a little skeptical, given how loose Uber is with the truth, but I see a possibility for them to succeed and smart people keep giving them good money so I want to give them the benefit of the doubt on the issue of actually making money eventually.
Depending on how large these economies of scale are, even the fact that a new entrants would be doing less volume (and therefore losing more) may not be enough to help them compete.