EPA numbers are standardized, there is no driver factor.
Yes, it's a known test that can and is often gamed by programming and eunuch tech so that the results are potentially closer to what your average dolt should get if they drive a similar cycle in a similar often unrealistically relaxed manner as the test. (Emissions testing is also gamed as displayed by the investigations and fines going out to Chrysler and almost all the trash German brands). Not every car is tested, either, there are formulas the OEMs can use to extrapolate data from platform mate vehicles (which is how you end up with CA suits, like Ford with the C-Max not meeting numbers). Manuals aren't as consistent as they are far more dependent on the driver, so ruling out that factor leaves them at a disadvantage, which I don't think the OEMs are bothered by as they don't
really want to sell manuals and they primarily remain in sports cars where people are less concerned about mileage. In the case of the twins, the ridiculously short gearing (to make up for the lack of torque, which is the true root of the issue) of the cars hurts them in testing vs wider-ranging CVTs and multi-speed autos in spite of their weight and cda being advantageous against other vehicles, like a Civic or—even worse with its less efficient AWD—a CRV. While I usually just beat the EPA highway ratings with a combined cycle when I drive (all previous cars were higher torque/lower hp engines with either outright taller gearing or the same top gear ratio only with 1/3 the hp, but 1/2 the torque peaking at 1/2 the rpm), I well exceed the highway rating on a combined cycle in this car. My point is that people are getting bad mileage because they are driving aggressively and don't care, are in heavy traffic almost all of the time, or don't know how to drive. These cars are capable of a lot better and it doesn't take hypermiling to do so. Hell, I got stuck in heavy traffic going to Cape Cod yesterday and was driving around 75 with several bouts to 100 on the way back and I'm still looking at 32 average on the computer (probably more like 30 when I calculate it at fillup). I keep beating this drum so that potential buyers don't think 27 mpg is the best this car will do if they would have to employ this as a daily driver and so settle for something else instead.
