Inside the gearbox of your EC-TEC drill is a one-way drive system. The primary purpose of this is to lock the chuck so you can tighten the bit in the chuck. It means that you have unrestricted power from the motor to the chuck, but no power can be transmitted backward from the chuck to the motor.
In the diagram below, the three green-colored pins and pink-colored 3-lobe coupler serve this function. When the motor is driving the output shaft, one of the gold-colored pins on the 3rd-stage gearset drives against the lobe of the pink coupler. However, the adjacent gold pin keeps the floating green pin centered in a small relief area.
If the output shaft turns faster than the 3rd-stage (gold) gearset, then the green pin is no longer held within this recess area and jams against the outer wall of the gearbox. This instantly locks the output shaft. (Important: This action is also why it is critical to never try to connect two drills together for a "torque-off". One of the drills will lock its output shaft and you will destroy the other drill.)
When the drill's battery is right within a small "sweet spot" of reduced power, the motor and output shaft can "coast" to a stop at the same speed and the green pins will not leave their recessed area. Another thing that can cause this is too much grease (or too sticky) in the gearbox. I learned this little fact when I reverse engineered the gearbox in the first place, and repacked all of the grease but put too much grease in the output area.

By the way, allowing this to happen too much, or at too high of a coasting speed may damage the EC-TEC electronics power transistors. This is because this type of motor will function as a 3-phase generator when you spin the output shaft. A little spin is OK, but too much will send power backward through the transistors. I have seen this happen
only when I had the motor and controller disconnected from a battery, so I suspect the EC-TEC controller has circuitry to protect against this that wasn't active without the battery.
Edit: My mistake. I suddenly remembered that way back when I first examined the C12 drill, I removed the 3 locking pins from my drill and then tried the "torque off" with another drill. This allowed the output shaft to be spun at high speed, which is what fried the EC-TEC controller.