This was kicked around last summer, I think, by some folks. The course code is very likely a variant on C, and could probably be hacked.

Also being a EE, albeit from the harware side, I can think of several resasons not to do it:
- Think about your responsibility to others as an engineer. You are looking to dork with a very tightly (and well!) integrated hardware-software-human interface which you are (probably?) not an expert in, nor have any particular knowledge of. Can you safely guarantee yourself and your passengers and the vehicles around you that you robustly found and regression tested every scenario? Things like 'just adjusting the setpoint for IMA' are rarely ever that simple...we all know that from hard experience. It's one thing to have a reasonable expectation of success when adjusting your own brakes, engine repairs, etc- the procedures are well-documented. There is no such analog for re-coding the ECU.

- IMA does kick in alone in eco mode on very light grades when the ECU decides it can handle the added load without the other 3 cylinders.

- ECO, strictly speaking, only means the car is getting over 30 (nobody really knows) mpg at that moment. 90% of the time, that means 3cyl operation, but not exclusively.

Having said this,

- If you happen to be a Honda programmer, my apologies.
- I and many others love the idea of a selectable 'constant load' CC as opposed to 'constant speed.' Like an amplifier that can be selected for constant power output or constant current consumption. IIRC, Ford said they were looking into this after the 'hybrid expo' day last fall.