These mpg metrics for electric cars are getting ridiculous

So how do these electric cars get huge miles per gallon without actually using gas? Cnet discusses the question. The bottom line is that they assume that no energy is used when the car is running on electricity.

The draft EPA methodology figures that a plug-in electric vehicle driver will go a certain number of miles on batteries alone and then another portion on the gasoline engine, explained Frank Weber, the global vehicle line executive for the Chevy Volt. To arrive at the mix between battery versus gasoline, the EPA is studying average American driving patterns, executives said.
The EPA is also developing another, less familiar metric for electric vehicles. In the Volt's case, it will take 25 kilowatt-hours to go 100 miles. Weber said the models behind the EPA methodology are "robust," adding that he expects the EPA to disclose more about the tests later this year.
To come up with 230 miles per gallon for city driving, GM assumes that Volt owners charge the car's batteries once a day, which enables them to do the majority of their driving from electricity drawn from the socket. The Volt, due late next year, is designed to run 40 miles on electric charge and then use a gasoline engine to sustain the battery for longer trips.
Misleading?
Triple digit combined fuel efficiency is certainly impressive--the Toyota Prius and Honda Insight hybrids both sport combined mileage of about 50 miles per gallon depending on driving style.
But immediately after GM's announcement, people began complaining that the claim is misleading.
For example, comparing the Volt to the Prius with that methodology is not useful, argued Darryl Siry, the former chief marketing officer at Tesla Motors and now a consultant with green-tech companies. By the EPA measure, it would appear that the Volt is many times more efficient than the Prius. But the 230 miles per gallon rating more accurately reflects how much gasoline has been consumed rather than the overall efficiency of the system, he said. "People will improperly conclude that the Volt is about five times more efficient that the Prius, which simply isn't true," Siry wrote.
A common way of doing comparisons is converting the embedded energy in gasoline to batteries, which a 2000 Department of Energy rule does in addition to considering the efficiency of the overall energy delivery system.
In the case of the Volt's city mileage, fuel economy will begin to drop off when drivers go beyond 40 miles before recharging. The Volt's electric driving range was chosen specifically because U.S. Department of Transportation research shows that almost 80 percent of Americans drive under 40 miles a day.
In the EPA model GM has followed, those first 40 miles equate to "infinite mileage," since it was charged from the grid and no gasoline was burned. But to consider electricity as infinite fuel efficiency can be misleading given that some energy--be it coal, natural gas, or nuclear--went into the delivery of electricity to charge the batteries. . . . .