While the value of state-of-the-art cars with V2V, V2I etc. is maximised with increasing numbers of other state-of-the-art cars o the road, it does not make them useless nonetheless even if it's just one car on the road. A smart car for instance is better equipped at a avoiding obstacles and crashes, optimising fuel consumption, keeping the people onboard connected with the rest of the world etc. etc. With such value, more and more people would buy such cars and the value increases further because of network effects - it's a virtuous circle.
I would venture to say that by 2050, most of the cars on our roads will be equipped with V2V, V2I etc. technologies and self-driving cars will be commonplace.
Yes Metcalfe has pointed about the value of a telecom/network infrastructure according to that the value of the infrastructure will be proportional to the square of number of users using it, but here the cautions to be considered are many more as it will be the multiplication of the square of the number of technologies/protocols and square of the number of systems in use.
On good thing that these self driven cars will do to all of us good drivers is that the Bad drivers will not be able to take us for a ride by their over speeding, lane cutting, honking, dangerously overtaking, tailing at high speeds and all that increase our BP and heart rate .
The thrill of driving one's own car goes away and will be restricted to the racing tracks and the cross country car rallies.
Oh I will love to see this scenario happening in my lifetime!