I think there is some value in at least one feathered landing during (certainly, initial endorsement) training, especially on prop aeroplanes with bigger engines.
There is no doubt that a successful feathered landing increases the pilots confidence. Rather like the feeling generated after one's first solo. The problem is one of cost benefit. The RAAF in the early 1950's found that to their cost after three Lincoln (four engine bomber) crashes during practice feathered landings where mishandling (pilot error) took place and the aircraft crashed and were destroyed during attempted low level go around after ballsed up asymmetric landings. Two of these were at Townsville in perfect weather conditions.
Case one was where the instructor was actually demonstrating an asymmetric landing with an outboard propeller feathered. He failed to detect the aircraft drifting to the side of the runway until too late during the hold-off and went around. Even with full corrective rudder he was unable to prevent the aircraft from yawing under asymmetric power and one wing hit a lone power pole on the grass between runways.
The aircraft crashed and caught fire but luckily the three crew were able to escape before the aircraft burst into flames.
A couple of years later an instructor was talking his student through an asymmetric landing with an outboard propeller feathered. The Lincoln bounced on touch down (and those tail draggers could really bounce). The instructor took control and tried to cushion the bounce by applying power to the three remaining engines. By now speed was low and the aircraft yawed sharply into the feathered engine as the instructor tried to let it down gently.
Directional control was lost and the aircraft crashed sideways and the landing gear collapsed. The three crew escaped before it caught fire.
Following that last accident, the RAAF banned all practice feathered landings on multi-engine types including the C130 Hercules and Dakota aircraft. I believe that rule still stands. Instead, for practice asymmetric landings, the engine failure was simulated by closing the throttle to zero thrust which approximated feathered drag.
The advantage of that was an all engines go-around could be made if the landing was stuffed up.
Tellingly, the then Department of Civil Aviation decided to ignore the experience of the RAAF accidents and practice feathered landings in civilian light twins continued in general aviation despite many accidents.
Finally:
and rejected takeoffs where the instructor would pull the power just before rotate speed
Leaving no room for error which means instant perfect reaction which most students are not capable of during initial twin endorsement. Risk mitigation was unheard of then?
the instructor decided to demonstrate the RTO procedure in the dutchess (can't remember what his thoughts were) and promptly blew a tyre and left us stranded near the intersection of two runways
Not at all surprising. Amazing what some instructors will do to inject a spot of realism..
