Popular Science this month looks at ways to make vehicles more fuel efficient.
The approaches are logical enough. Reduce aerodynamic drag, reduce weight, recover energy lost to heat in braking, and recover energy lost to heat in exhaust products. But, by and large, they are not being done.
At first, your intuition is assume that this is because they can't be done. But, the Toyota Prius hybrid, which owes much of its efficiency to low drag, low weight and recovery of energy lost in braking, dispels this idea. As the article notes, industry hasn't exactly been trying hard to address fuel efficiency (emphasis added):
The average price of a gallon of gas is higher than at any time since the early 1980s. The Middle East seems more volatile than ever. And even climate skeptics are starting to admit that the carbon we’re pumping into the atmosphere might have disastrous consequences. To these circumstances, automakers have responded with a fleet of cars that averages 21 miles per gallon, about four miles per gallon worse than the Model T.
The energy recovery from braking and exhaust heat could both be almost invisible to the driver, and hence, these are prime choices for widespread adoption. And, the approaches can both be done in the same vehicle without interfering with each other materially. This doesn't have to be an either/or technology choice.
It is hard to know whether to feel elated that there is some hope for big improvements in fuel efficiency, or to feel betrayed that the industry has never made a serious effort to address the problem before, as none of this is really ground breaking technological development in concept, even if it could have immense practical importance.