The laminations are pressed onto the shaft like all rotors. The glass tape is applied to the green armature and then it is treated with a polyester resin under controlled conditions of temperature and such. The result is a thick layer with deep penetration of a very strong and high temperature encapsulation/insulation. This process has been used since the 1970/80's and eliminated the top slot stick armature assembly method. On an armature this size it would probably hold to 10kRPM but the comm and balance putty would likely hold the recommended speed to 5k.
Each of the copper bars you see in the photo is an armature coil side. I'm thinking this particular armature has 37 of these armature coils, sometimes called hairpins due to their shape before they are inserted into the core. Each of the 37 armature coils has a single turn, or 1 T/c. There are 37 segments (or bars) on the commutator and 37 slots in the lamination core. Each slot has 2 coils sides one on top of the other. Each of these 2 coil sides belongs to an armature coil of opposite polarity, so to speak. So, these 2 coil sides in the common core slot must be insulated from each other. If that insulation fails, there exists a "shorted" armature coil. It may in fact actually be a pair of shorted coils, but the result is the same.
When that short first occurred, it was likely at high speed/high voltage and there was arcing and burning of the surrounding area. It could have resulted in a catastrophic failure or a lesser of an event even allowing the user to get home. But then jas notices this 1/4 turn behavior. And that is indicative of a shorted armature coil attempting to be rotated thru a 4 pole energized field. Shorted coils do not like to move in the presence of magnetic fields.