Americans traditionally measure the fuel economy of their cars in miles per gallon. Europeans tend to use liters/100 km. Americans not only use their customary unit rather than metric, but also have a measurement that is reciprocal to the European one. (To convert one to the other, divide 235.215 by the number you already have – although official stats may still not be comparable, because the U.S. and Europe use different test-routines to determine city/highway mileage. FWIW, I’ve heard that both U.S. and European tests are too optimistic about fuel economy compared to the real world, but European tests are a little more optimistic and further from reality.)
I think metric is better in general (but Europeans really should have done liters/1000 km! Then their measurements would have two digits of precision for typical modern sedans, crossovers, and such without going past a decimal point, plus 10^6 m is a better metric distance standard than 10^5 m.), but fuel per distance and distance per fuel both have arguments in their favor: do you want to know how far you can go on a tank of gas (American), or how much gas you’ll need to go a given distance (European)? Is it better to have bigger numbers be better, so good fuel economy is viscerally impressive in the same way as high horsepower (American), or is it better to have a number directly proportional to fuel consumption, so it’s obvious that going from 12 L/100km to 6 L/100km is twice as important as going from 6L/100km to 3L/100 km, whereas in the American system it’s less readily apparent that going from 19 MPG to 38 MPG saves twice as much fuel as going from 38 to 76 MPG? (assuming constant distances) I think the European system of volume over distance is probably somewhat more sensible on the whole, but not as clearcut as the metric vs customary issue.
What strikes me as interesting, though, is to consider the European system: liters per 100 km. That’s volume – x^3 – over distance – x. Phrased like that, you can see the immediate temptation: what if we reduce x^3/x to x^2? Now our units of fuel economy are in area. If we have a vehicle that uses 6 L/100 Km, we can phrase that as 6 * 10^-3 m^3 / 10^5 m, which is 6 * 10^-8 m^2, about 6% of a square millimeter.
Does that actually mean anything, though, or is it just an artifact of misused dimensional analysis, like claiming that torque can be measured in joules because newtons * meters = joules, when really torque just isn’t energy? I think the area measurement of fuel economy actually does have a physical interpretation: if you were driving, and a tiny filament of liquid gasoline with this cross-sectional area traced your route, then that filament of gasoline would be approximately enough fuel for your trip.
I can’t think of a physically intuitive explanation of what the dimensionally-reduced American measure of efficiency (reciprocal area) is, though.
