Originally Posted by Hypnotoad
The EPA would treat it the same as a manual transmission. When a car maker performs mileage testing with a manual transmission to present to the EPA they have a professional driver trained to shift for the best possible mileage. The EPA knows that not everyone will shift for the best possible mileage but they accept the best-case results. They would certainly do the same for a "trainable" automatic.
No wonder some owners have trouble getting mileage as good as testers do for the EPA numbers.
I thought the DCT is already "trained to shift for the best possible mileage" if left on its own without driver intervention. Do you know whether testers used the select shift feature at all in tests for the current mpg results?
I'm not as certain as you are that the EPA would be willing to ignore the possible fuel economy impact of a "teach mode" like what spektrum2
has suggested, but I'm just speculating. If you have insider knowledge of how the EPA or carmakers work, I'll take your word for it. If they would ignore it, they could just skip testing and stick with the current DCT test results.
After giving more thought to the "teach mode", I think that even if adding it improved
fuel economy, Ford would not want to until after they get their current problems resolved. Adding complexity with any new user (previously known as "driver") involvement would make that harder. Those pesky "users" always find ways to mess stuff up and give software designers nightmares. ("We never imagined the users would do THAT! Tell them to not do that.")