Using a strict nrg to equivelant body mass probably isn't a good idea (hugely wasteful). I would do nrg to equivelant shell mass, I think. That's like 10 to 1 compared with body, right?
Angular momentum is broken presently, because it wasn't based on proper laws of motion. If you fixed up the behind the scenes stuff to incorporate Moment of Inertia and Torque, though, aimdx, etc. could be modelled as motors: they produce as much torque as needed to achieve the desired change in angle, and charge the bot the equivelant nrg.