I'm running an MP II, which requires an external controller. I'm not using a GM external controller and am not familiar with the shunt modification that Bikemad made to his, although the principal of the thing is easy enough to understand.
Changing of the shunt's R value looks to the controller like more or less current is being passed. It's a simple cheat really. Typically, one would lower the R value to fool the wee brain of the controller into allowing more current to pass. The lower R value produces a lower voltage drop across the shunt. The lower drop looks like less current and so the controller turns the FETs on longer to compensate. The problem then becomes one of saturating the FETs and so you double up on those. Pretty soon you have a new business making high powered bicycle controllers.
I haven't had a look at the windings inside my motor. Having said that, I wouldn't be shy about changing the Wye/Delta configuration, if there was reason to. I would rather work on an MP II though, which
must have more free space inside than an MP III, which is packed with the internal controller. In either case, it's a matter of identifying the ends of the windings and changing the way they're hooked up together.
I would expect an internal controller to be a bit more of a pain to work on because the windings are (probably) soldered directly into the controller board. Not sure what they're using to connect the windings in the MP II, but they might even be using simple Marr connectors. I'm sure Bikemad can give you the details.
In your case, you'd want to convert a 700C Wye (standard configuration) to Delta (higher speed, lower torque). The only real trick is making sure your Hall Effect (position) feedback sensors are in proper sync with the windings.
