The current version of the car is not what the technology will become. To judge the final impact of driverless cars on the first version being made as proof of concept is ridiculous.
This article is what would have happened if someone had reviewed the Benz Patent Autowagon and projected it into the future traffic needs of today. Yes it would have been terrible but it made today's cars possible.
Driverless cars when significant will not only not wait at traffic lights. They won't need traffic lights. Does Google know this? I don't know but they will.
Why ask what happens in tunnels with GPS? Do you make up stories about technology without researching the current state of consumer GPS devices?
The better GPS navigation devices use accelerometers to fill in when the GPS signal is unavailable. Phone based GPS systems of course use accelerometers and other signals to fill in when GPS is unavailable. Beacon technologies are becoming standard in phone systems to provide indoor location more accurately.
It would be quite easy to put beacons on the exits in tunnels thereby providing quite accurate navigation within the tunnel.
The car has LIDAR so it can tell what is happening inside the tunnel - and could take the appropriate exit the same way a human does without a GPS.
If you are going to make commenta about this car at least make ones that are not obvious garbage.
I by no means think this car is in itself the big thing of the future. The technology it introduces to the market is the start of something big though.
Im not even a fan of Google. Just really would prefer not to be sharing the road with frustrated and somewhat incompetent humans and to move to a time when traffic lights are gone. Traffic lights waste time and petrol. Won't miss them.