I think your analysis it on point except for one part: the barrier to entry for becoming a driver is actually non-trivial and perhaps more importantly there is

I think your analysis it on point except for one part: the barrier to entry for becoming a driver is actually non-trivial and perhaps more importantly there is not a large supply of drivers in a given region. The pool of people who are looking for a new job, do not mind being a taxi and have a suitable car is somewhat small, and this is why uber and lyft must campaign heavily to recruit these folks. If this pool of people took collective action it would be very hard for Uber to come in and replace them - they could likely flip a single region to a new coop if they could self-organize. That said the fact it could happen it not really related to if it should happen - as you say it's unclear how much extra money would actually flow to drivers as potentially Uber is operating on thin margins.