I’m no Tesla fanboy, but franchise models exist now, why couldn’t it work for cars?
Leasing cars to Uber drivers has taken off in the last decade. No reason for Tesla to sit on a quickly depreciating asset like a car if they can double dip (initial purchase and Taxi profit share) without having to pay for electricity, depreciation, or maintenance.
An Uber driver is in their car while the car is "working". They have the ability to react to any situation because they're physically present. They can protect the asset they have leased.
A Tesla performing overnight robotaxi service while the owner sleeps is a liability to the owner. The car may be damaged, abused, or vandalized. This may be done by people other than the person who requested the ride. The car can notify the owner, but now the owner has to get there (without their car, because their car is there) to react to whatever is happening, long after it has happened.
The proceeds would have to be absolutely massive for me to put up with that. I don't want to run a one car cab company. Offloading the ride hailing and driving aspects is nice, but what's left for me is not a job I want to do. I don't want to wake up at 3 AM to a biohazard notification from my car.
It's just not a good deal for the owner. They're eating all the negatives you called out and taking on substantial risks.
It might be possible to have a robotaxi fleet that makes enough money to pay for the staff and absorb the risks. Losing one vehicle is not a big deal to a fleet. Losing my vehicle is a huge deal to me.
"Your car is also a robotaxi!" is a gimmick for those who have trouble connecting events over time. There will be those who gleefully shout about every robotaxi revenue deposit they get while ignoring what they spent to get it, like a lottery scratch-off player with a $100 win after weeks of losing tickets.
The Tesla robotaxi service might fail for a variety of reasons but I'm not convinced that this is one of them. Turo has already somewhat proven the business model of private owners renting out their cars. You can find occasional horror stories of cars being destroyed, but car owners keep using the service.
Turo is a pretty different product than ridesharing. Much closer to traditional rental cars than robotaxis.
For fun, I went to Turo's site to see the shortest interval I could rent - looks like 1 hour. Prices are day-based, like rental cars, and thus are pretty expensive vs what you'd expect to pay for a rideshare. They have very different use cases.
The liability questions are clearer for Turo - you are renting the vehicle to someone for a set period of time and they're responsible for the operation and condition of the vehicle while it's in their possession. That looks completely different in the context of a rideshare customer. The customer is not in control of the vehicle. The customer is only accountable for their own actions.
A robotaxi is going to do rideshare-like behaviors. It will be picking people up and dropping them off from popular spots, and doing it often. A rented car from Turo or anywhere else will have different patterns of usage, many of which will look pretty similar to someone just going about their day. This means robotaxis are easier to identify and target.
Unfortunately, I can't remember the source, but I've read that Waymo has at least broached the idea of entities other than Waymo owning vehicles, as in a franchise. They worded it somewhat awkwardly, referring to "private ownership." which some people took to mean that they could own their own Waymo, like a private car, but I'm pretty sure it was intended as a business model variant that could support faster build out.
The economies of scale of running a large scale operation would make it much more likely to reap a profit worth pursuing. In addition, there's no way Tesla wouldn't be directly impacted by the reputation of a fleet of privately owned vehicles running their software but in various states of repair and cleanliness. If it's actually the panacea for local transportation it's sometimes made out to be, it'd make no sense for Tesla not to run it themselves and have full control.
If the vehicle is solely used for a taxi. I don't see how it works if you are setting a window of availability. What happens when traffic gets so bad it can't make it back in time. What if it gets in an accident? One of the main reasons I have a car is for convenience. If my kid gets sick at school, how long before my car comes back?
Leasing cars to Uber drivers has taken off in the last decade. No reason for Tesla to sit on a quickly depreciating asset like a car if they can double dip (initial purchase and Taxi profit share) without having to pay for electricity, depreciation, or maintenance.