The simplest way to increase overall fuel economy(assuming about 40% highway and 60% city driving, and assuming driving style remains constant) is to do the following, in order of importance:

1) Reduce aerodynamic drag
2) Reduce weight
3) Reduce engine speed at cruising

There are people at ecomodder.com, gassavers.com, and other places that have doubled the fuel economy of their cars doing these things plus adjusting driving technique.

The automakers have repeatedly designed no-compromises 60+ mpg midsize and/or high performance sports cars as far back as the 1970s that ended up remaining concepts, even when market studies indicated that there was sufficient demand to make a profit on them.

Many of these, like the 2002 Opel Eco Speedster(turbodiesel, 94 mpg US combined, 160 mph top speed), are streamlined and lighter variants of already existing cars.

I sincerely hope that Mazda and others start offering low drag cars to the public. A diesel sedan with an inline-6 3L turbodiesel making about 300 horse could easily achieve 50+ mpg if it was in a chassis that had a 0.16 Cd, 20 sq ft frontal area, and used construction methods similar to a Kia or Hyundai to save weight, keeping it at a svelt < 2,800 lbs. And imagine the acceleration...

The automakers aren't too enthusiastic about load reduction though; it dramatically reduces maintenance by putting less stress on the components. This is why the 240D Mercedes is so reknown; while it didn't use load reduction as a strategy, the components were highly overbuilt, and taking a conventional car and streamling/lightening it, without changing to lighter duty mechanical components, would produce similar reliability.

I disagree that electrics are impractical. High school kids have managed to take Chevy S10s in the 1990s and make vehicles comparable in range and better in performance than the ones the OEMs were leasing out at the time, using off the shelf components(eg. Seth Murray's electric S10).

There are electric pickup trucks using golf cart batteries that have been able to exceed 100 miles per charge at highway speeds of 70+ mph, and do 0-60 mph in < 18 seconds(see Tony Ascrizzi's "Red Beastie, a Toyota XtraCab with 120 miles range).

A company called solectria, founded by James Worden, built a car called the Sunrise. In 1996, it broke a world record for range that hasn't been beaten today, going 373 miles on a charge. This was a midsized car that could seat 5. Granted, James used hypermiling techniques to increase his range, but this car had a real 200+ miles range when driven agressively on battery technology that has only one-third the specific capacity of today's lithium ion batteries. How? It addressed aerodynamics and weight. GM and other major automakers refused to touch his design; it would have been $20,000 in mass production, had no transmission and could still reach 75 mph, and had a real-world driving range of 200 miles per charge or more. An oil company has since acquired rights to large format NiMH batteries > 10 AH, and this chemistry does not take too well to being charged in parallel, making large format the only practical way to use NiMH. Being that large format NiMH using Stanly Ovshinski's technology is kept unavailable by big oil, small companies making hand-built EVs have to look to less reliable and more expensive to use batteries like lithium ion(although they offer better horsepower and range than NiMH) for their EVs.

Plug-in hybrids like the Volt would actually be more expensive than pure EVs to operate. Why? Wheras pure EVs with 150+ miles range will rarely see their battery deep discharged, plug-in hybrids with only 40 miles range will see deep discharges as a mater of routine. Deep cycling is extremely stressful on a battery and will cause it to fail more quickly. Thus, GM's Volt has a pack sized for 80 miles range, with only 40 miles accessable, and it is a very expensive battery on a per kWh basis, even in mass production. A battery that lasts 500 cycles to 100% discharge in a car with 250 miles range will last 125,000 miles minimum. That same battery in a car with 40 miles range will give out at 20,000 miles, hence why the Volt needs a more expensive technology than lithium cobalt, driving prices up further.

Tesla Motors, on the other hand, can use conventional lithium cobalt technology and still get appreciable battery life; while the battery technology in the Tesla Roadster has dropped to $400/kWh for the components, since the car is virtually hand-built(along with the BMS, thermal management, ect.), the cost to implement it in the car is about $700/kWh. Mass production of EVs would change this, but Tesla does not yet have mass production capability. They are working on it, and know that an affordable, mass market, 150+ mile range EV is possible; it's been possible for almost 15 years(or 35 years if you were willing to settle for the 100 mile range limitation of the 1970s using golf cart batteries in cars like the CDA Towncar or the EXAR-1).

Further, the Volt is an unaerodynamic, heavy pig. It weighs about as much as my 300 SDL, and has a Cd*A that is par for the course when compared to what the rest of the automakers are offering for the 2010 model year. This makes its battery cost per mile of range increase dramatically.