I have read many times about the magnets loosing power when the rotors are removed, however I have not observed this to be true. I found an old Kollmorgen in the junk, it's rotor had been removed. I took the parts home and it sat on a shelve for almost a year before I put it back together.
Once it was together it spun freely by hand, shorting any two leads makes it very difficult to turn, short all three and it's almost impossible to turn.
Turning the shaft by hand and touching any of the main power leads together and you'll get a small spark.
Granted, I have no way to compare it to "before" but I can't imagine that were would have been a major difference.
I've also taken apart a couple of AB servos and haven't noticed any difference between "before and after" (using the shorted leads method).
I have also taken apart an Emerson servo and had the same out come.
I've taken apart more than a dozen (junked) servos and each one of them had the magnets on a magnetic rotor (keeper) and I still don't understand why or how these magnets would lose their magnetism, whether the rotors are removed for moments, hours, or weeks.
Now for my disclaimer!! I have never taken apart a servo, with out trashing the encoder/tach/reslover. Thus never verified it's level of performance after it's been dis and re- assembled. Maybe the magnets did lose something, but I doubt it was a whole lot.
I would NEVER suggest taking apart a WORKING servo motor, unless it's your intention to "trash" the "servo" part of the motor.
Again, I've seen many posts saying that removing the rotors will permanently damage the magnets So I guess it must be true, but I haven't been able to verify this.
I do believe that taking most any servo motor apart (with out the proper tools/calibrations) will permanently wreck it's ability to perform as a SERVO MOTOR (but not necessarily it's ability to act as an alternator), and often have wondered if this is why people say that the magnets will lose power.