That's not even a 20% improvement. As easy as it is to manage with the new ECU's its probably worth it for the across the board improvements in economy, but individually it would be a lot of work to engineer.
If you can get a consistent 5% fuel economy improvement than you have just struck gold, so if somebody came up with a way to improve economy by 20% they would be buying their own private island within a year. In my work we also perform all types of fuel economy testing and a 2-3% improvement is considered a success.