The argument of the 10-part series they put out on Uber is that main expense is the actual transportation, and you can't achieve meaningful economies of scale, which is the reason that livery services never expanded beyond regional monopolies previously. Relatively speaking the software is a commodity (and traditional taxi services remain competitors to Uber anyway -- likely more attractive in many cases if you take away the subsidized price difference).

Yes, that was more or less my point.

And also Uber (more than Lyft) it's by far the major player, it's entrenched in people's imagination as the representative of the very service it provides, and it just works.

"Call an Uber" is as idiomatic today as "Google something".