> Given that cars 95% of the time are not moving anywhere, then self driving cars could reduce the car consumption by 10 times, an order of magnitude less.
I question this assertion.
Granted, I'm coming at this from a semi-rural American perspective, but I don't think true full self-driving would cause a reduction in emissions. If anything, I think it might increase them.
Urban areas are where emissions are most immediately an issue. Those areas also have limited parking. If vehicles were capable of autonomous operation, I would expect more people to drive, exit the vehicle at their destination, then allow the vehicle to park itself. This would incentivize companies to add additional offsite parking, which in turn would mean that vehicles would drive themselves to and from those parking areas - a net increase in distance driven, and therefore emissions.
That's not even considering things like vacations. My family and I drive 10+ hours about once each month. About half of those trips are for pleasure, and the only reason we take one vehicle is because it doesn't make sense for both my wife and I to drive separately. Otherwise, if I'm spending a week at the beach, I'd prefer to be able to drive my Jeep around while we're there. My Jeep doesn't fit my family and our luggage comfortably, so we don't take it right now. If it could drive itself, the only reason not to have it come along would be the cost of fuel and wear.