I believe mass car transport has a lot of problems, where the tailpipe exhaust emissions is only one of them. While electric cars does indeed solve the latter, I do agree that buying a Tesla in order to be "green" hardly make any sense.

Another problem that seems to be soon resolved is the dependency of an human driver to concentrate on the task of driving. This is a safety hazard (as humans are generally bad at driving), a hell of a lot of human-hours are spent behind the wheel, and since it's a big insurance cost involved when renting cars, that's also quite expensive. Well, some people actually like driving, they will probably not be able to see my point.

One thing that will happen once we get truly autonomous self-driving cars is that the cost of taxi transportation will fall a lot - after all the driver is the driving cost of using a taxi. With this, private car ownership will decline - there is not that much of a point to have a private car if one can get a car cheaply within few minutes simply by pressing a button. It's even better, as one can order a car with big luggage space or a car with space for 8 people or whatever one needs ... rather than having to do daily commutes to work with a car designed for going for holidays with the full family.

This doesn't mean Tesla is a scam, it may very well happen that Tesla will get a dominant position in this market, after all Tesla is one of the major players when it comes to research into autonomous vehicles.

In my opinion, we could have had networks of electric self-driving vehicles some ten years ago if there had been a strong political will for it; there were already experiments done with self-driving cars in the 80s, and there are dozens upon dozens of well-researched ideas for autonomous vehicles. If one has a separate right-of-way for autonomous vehicles, has well-working protocols for vehicle-to-vehicle-communication, infrastructure meant for machine-to-machine communication rather than road signs and traffic lights meant to be seen and understood by humans, autonomous driving is trivial. Once it's a requirement that the autonomous vehicle should navigate in mixed traffic and read and interpret road signs meant for humans, "autonomous driving" gets into a very hard AI-problem. I'm actually very surprised that self-driving cars has gotten as far as they have.

The problem is mostly with inertia, we do have a well-established road network, we do have a lot of regulations detailing what is a "car" and what is not (and if something doesn't follow the regulations and cannot be classified as a "car", "bike", "motorbike", etc, it simply cannot use the public roads), we do have a lot of car-owners who love their cars, there is a lot of staked economic interest in keeping status-quo, etc. Hence I think it's unlikely to see the cars go away in the near future.

Ten years ago, I would have expected some hybrid technology to come to the rescue, vehicles that would be fully autonomous when driving on infrastructure built for it, but could be used as an ordinary car on the roads.