Discussion on:

A GPS that can choose the most energy-efficient route?

It would have to use different algorithms for electric and gasoline, I would think, as stop-and-go might not affect them the same.

More than that, until surface street traffic and traffic signals are taken into account, it isn't going to be very effective. I have a GPS device with traffic, and it sometimes suggests using the access (service/frontage) road when the freeway is backed up, not aware of the backup on the access road. And once off of the highway and on city streets, it severely underestimates the effect of traffic signals. Neither the FM/RDS system, nor the new HD (Garmin calls it "3D") traffic provide surface street information.

Google traffic does provide the surface data, so perhaps we aren't too far off.

If its not going to take into account the driver's style, I don't know how much help it'll be. For instance I've heard that most cars get the best mpg at 60 mph. But most people on an empty interstate are easily going 75-80. An algorithm that assumes that drivers are going to drive optimally is doomed to fail...or lie to drivers.