And they've gotten used to the city/highway/combined figures. (Where "combined" is largest.)
I think the "
http://www.epa.gov/EPA-IMPACT/2000/June/Day-12/i14446.htm">Petroleum Equivalent Fuel Economy" method works great, to put things into perspective, except for the fact that (in the case of of the "Volt") there is such a radical shift in efficiency when the mode of operation changes (from "pure EV" to "Extended Range EV.")
The EPA method GM used makes sense. I am disappointed that they (apparently) chose not to include the PEFE calculations for the electricity.
Here's a method which just occurred to me. All cars made in the last 10 years have had OBD-II interfaces. In New York State, annual automobile inspections use the OBD-II data (if the car is new enough) for emissions testing. The data actually goes "off to Albany" as part of the inspection process. Every year, data is extracted from every "new" automobile, and most of the "old" ones as well.
It should be possible at that time to determine the "EPA Rating" for each individual car, the data could be compiled by make and model. This would allow a "real world" figure for efficiency. (On average, drivers of this model got this sort of efficiency.)
Unfortunately, figures would have a 1 year lag, meaning that for the first year of sales figures for a radical new design would be approximate (your mileage may vary) but after that first year, the EPA (and customers) should have a very good "real world" figure for the "average driver."