The point I was making wasn't just that it's not feasible, but also that it is counterproductive to what we should be doing - reducing the number of cars on the road. Why burn money on testing infrastructure that will not be needed down the line when (hopefully) less people need cars?