I don't see why meta-Uber has to hitch on the backs of Uber and Lyft. Just have an app (I'd suggest "Ride") that knows all these local co-ops and contacts an appropriate one. Ride Austin when you're in Austin, Ride Shanghai when you're there. Make it trivial for anyone to start such a co-op in a new city, and the drivers will come. There's no need to first book on Uber.
This.
If the co-ops exist, the meta-app is not a difficult app to build.
Geolocation to match you with the local collective and an API to send the booking request to them (unlike Uber they want to work with metasearch companies). Done. Let the collectives handle the certification, pricing and local routing.
(I know someone who founded a business effectively doing meta-search for a ride with local minicab companies via nationwide telephone enquiries, so it's not like the basic idea wasn't possible even with yesterday's technology).
Trouble is, for the most part these cooperatives don't exist. Which makes the meta-app less useful than Uber in most parts of the world. Uber has every incentive to discourage people from using a meta-app as their first port of call for ridesearch, and so even if their TOC were completely unenforceable they'd just find ways of blocking a meta-app from accessing Uber data. Blocking people and drivers who are obviously hacking Uber's service to co-locate each other, not paying for the ride via Uber but both ending up in the destination anyway is even easier. A non-profit meta-app isn't going to win an arms race with a multibillion dollar enterprise whose business depends on being a more reliable way of booking rides. Legal restrictions on reverse engineering their software isn't their moat.
The problem is that Austin is an edge case. Drivers got together and formed it and raised money to build the app and set up a vetting process the city govt approved because they'd lose their livelihood if it didn't exist. RideAustin doesn't exist in most other cities because drivers and passengers aren't really set up for collective action, particularly not against the company sending them rides, the costs of getting the service working aren't trivial and any nascent startup would have the problem that Uber was more recognised (and would likely try to kill their app stone dead by subsidising Uber rides with VC money). If you want more local cooperatives, you need to bring down the software development and vetting costs to run a service like RideAustin, not reverse engineer Uber's iPhone app. Or lobby more local governments to find reasons to bar Uber and Lyft from more city centres. And the market will see more cooperatives happen organically if/when Uber tries to squeeze more profit out of rides...
The reason why Cory wants to hitch on Uber is because he knows people trust Uber more than a random app they've never heard; he says so himself, when he recounts his Shanghai experience. So you'd find the driver on Uber (borrowing its trust network) before switching to the other app. But that also shows why his argument against Uber is flawed: they can charge more money because they actually provide a better service by being available worldwide, unlike these local apps. If the Meta-Uber were to kill the host, it would destroy itself.
But people who don't trust the other app are not going to use the other app anyway. If you want to hire a co-op driver, start with the co-op app first. Hoping Uber will accidentally send you a driver who also happens to be in the co-op seems like a pointless exercise.
As a rider, I need to trust the app to do two things: give me a decent driver and provide me with a reasonable estimate of the cost of the trip. By first making the reservation with Uber, I can get both of those from them, making me much more willing to use the local app to do the actual transaction.
