> If the efficiency of an electric car is about 35% based on the electrical
> energy used to charge the batteries, and the efficiency with which that
> electrical energy is produced and delivered is also about 35%, then the
> overall efficiency of the car based on the heat produced by the fuel
> burned at the power plant would be only about 12%. That may be the
> fairest basis for comparing the efficiency of a gasoline car with that of
> an electric car.
No, no, no. Using electrical cars has primarily one function: zero emission, at least in situ. Properly designed electrical cars should be lithium-cell driven composite-frame lightweight vehicles without a transmission (motors in the wheel hub), having a spike cache (e.g. a supercapacitor bank) and regen (brake energy regeneration) braking. These things would be environmentally clean, silent and have very impressive driving characteristics while being impact safe. Design studies of these have been made.
Of course the most sensible thing would be using a methanol (reformer) fuel cell or a hydrogen (metal hydrid/buckytube adsorber in composite pressurized tank) for energy source. The hydrogen could come from fossil fuel reformers or electrolysers (which can be pretty efficient nowadays).
There are lots of sensible alternatives to the way we're usually doing things. Unfortunately, the general public seems to have teflon lungs and glass eyes, and don't mind the beautiful St. Andreas mountain view be obscured by the veil of photosmog.
Of course these R&D expenses should be better put into development of molecular devices which could run rings around any Carnot or fuel cells, but the point is R&D resource distribution is not a zero-sum game. We are very, very wasteful.