It refers to the eddy currents. Some kinetic energy is lost when the field around a magnet builds up the field lines in a nearby metal, and you don't get the energy back if you move the magnet around. I personally think of it as little "cells" in the metal that contain arrows. The magnet makes all the arrows line up. Once the magnet is moved, the arrows spin around either to a new direction or randomly if the magnet is taken away. Not sure if that's the right way to picture it, but at least it helps imagine the "eddies" spinning around. It generates heat. An eddy-current brake is designed to do this very effectively.
The eddy current loss happens in all metals, steel, iron, copper etc. but in a motor like the one you are converting, the magnets are interacting with the laminated iron stator. If you try to measure a torque from iron loss (pretty hard on a small motor) it's really hard to separate out the iron loss, the friction in the bearings, and some kind of "average" that overcomes the cogging, so they are sometimes all lumped together when we discuss it.
You should try googling the high-school science demonstrations where the teacher drops a ball bearing down a copper tube, then drops a magnet down the same tube... takes a while for the magnet to come out the bottom.