It has been often said that there are lies and then there are statistics. Put another way figures don’t lie but “Liars figger”. Yet it often comes down to scouting down to the methodology and methodologies employed to check how a given set stats , specs or figures are derived. To look under the hood so to speak. Yet few do this – and many are taken to wrong – misguided decisions. In this case its checking how claims of increases of mpg - that is miles per gallon driven in a vehicle are derived.
In this case the numbers and calculations seem fair and honestly derived. Researchers at the U of M took a sales weighted / EPA mpg approach. That is they took the EPA rating of a given vehicle incorporated its monthly sales and went across the board with their calculations.
In most markets and indeed life its a 80 /20 rule that is 80 percent of anything comes from 20 %. Hence by covering the most popular models most of the volume of auto sales in America in a given time period – month , year or a number of years is covered.
True the numbers may not add to fully 100 % but for the most part its more than enough of a sample to do the coverage.
Yet remember your own circumstances and situation. For example if you tow a trailer behind your truck a larger engine may well be more fuel efficient and also less wear on tear on engine components than if the vehicle was unencumbered .
more can be read at this link: http://www.autoblog.com/2013/09/12/average-new-car-fuel-economy-record-24-9-mpg/