Maybe. It all depends what is limiting the current. There are several factors. By default, I think Colin's instructions call for setting the reference voltage on the driver to .39v. This equates to a current limit of 2.5*.39 = .975A (for a Pololu clone board). The motors are rated for .5A. This is one of the reasons they get so hot. It's more complex, though, because the resistance of the coil is 19 ohms, so at 12V, the resistance will limit the current to .63A, which is why the motors don't actually burn up. It also means that if you have your driver current set above .63A (ref voltage of .25v) I don't think you'll see an actual current increase. (Microstepping makes this more complicated.)

The killer, though, is the inductance. 32mH is really high for a reprap motor. This won't affect holding torque when nothing is moving, but it limits how fast the current can rise in the coil at a given voltage. As you step faster, the time for the current to rise gets shorter and at some speed, it can't rise enough before the next step and you start losing torque. At some point, the motor can't overcome the load and you skip a step (or twenty).

I think this is what is happening during the retract.

Most of the time, Slic3r is configured to retract 1 or 2 millimeters when moving between islands. It should quickly retract, move and then quickly advance the same distance it retracted and start printing again. You should see the gear turn backwards some distance, and then move the same distance forward when printing resumes. In my case, it was stuttering and clicking as it retracted, and then running smoothly when re-advancing, creating a blob. When I slowed it down to 10mm/s, it started retracting smoothly and re-advancing the same distance. (There may be something here involving nonlinear loading and gear backlash, bit slowing it down seems to help.)

If you print something with gaps that the nozzle has to cross (like the oozebane object) you can see the retracts. Try it at 10mm/s and watch how the gear moves. Then turn the speed back up and see if it's missing steps.

My Itty Bitty Belted Extruder (http://www.thingiverse.com/thing:389105) uses a lower gear ratio (4:1 vs about 5:1 for the Greg's) and it uses a NEMA 14 motor with about 1/3 the torque of the motor that Colin ships, yet it's able to easily retract at three times the rate without issue. So I don't think it's a torque (or current) issue. I think it's the motor winding inductance that's limiting the step rate.

Keep in mind that I'm entirely self-taught here, so if someone with more experience or formal education wants to shed light on this, I'm all ears.