A Duke University study, published in Science and reported in the Consumer Reports, found that miles per gallon are misleading and that consumers would make better decisions if fuel consumption were posted in gallons per mile.
Well, duh. Shouldn't that be obvious?
Is it any surprise that people calculate better with direct quantities than with inverses? Miles per gallon is an inverse measure of the cost of fuel, so one doesn't have to be a rocket scientist to figure out that consumers will have more trouble making rational decisions based on that measure than based on a direct measure of fuel consumption, such as liters per 100 kilometers, which is used in Europe and pretty much the rest of the world.
Most people don't realize (at least not immediately) that driving 45 mph from Startington to Endington (distance: 60 miles) and 80 mph on the way back takes more time than driving 60 mph both ways. But they do realize that a 80-minute trip and a 45-minute trip together take longer than a two-hour trip. Similarly, people don't realize that going from 12 to 16 mpg is a greater saving than going from 24 to 32 mpg, but they would recognize immediately that going from 20 to 15 l/100 km is more significant than going from 10 to 7.5 l/100 km. Elementary, my dear Watson!