What Karbine said is correct. I myself have attempted to create an E2 that applies both forward force and counter-torque based on RPM, to more accurately simulate a propeller. I may release the code. It went something like this:
Gotta love that dirty code.
I'd still have to work out the proper forces to figure out how much force would realistically be output per amount of torque (which involves figuring out how much weight a typical engine of x power output can move and other things to give me a power curve that I can reference to figure out how much force needs to be applied depending on the amount of counter-torque applied and vice-versa), but that requires maths, and I hate maths.