The key issue here is if the tail rotor thrust line is above the center of gravity and then we unload the main rotor for some time, not a fraction of a second, but a little bit more, a right turn will develop. There is a long tail rotor leverage.
This is more porne to happen if the speed is high as the nose will be low and the tail high.
If the negative G situation is kept long and hard enough, it is possible to strike the tail cone. The main blades will for sure flex enough.
A hard landing, for instance, this is with a high descending speed will possibly make the blades to strike the tailcone as they unload.
The next problem is when we unload the main rotor at high speed and we experience the right turn is to try to counteract the turn by just moving the cyclic to the left. The unloaded rotor will allow the rotor head to move freely without "pulling" the helicopter, remember, no thrust situation, and the rotor head (actually the blade roots), will hit the main shaft. If it is hard enough, they will shear the shaft...
So it is a must to first load the main rotor so it can "pull" de helicopter, and then correct the turn with the cyclic.
The R22 has a very, very quick and noticeable right turn tendency at high speed, real quick but at low speed, say 40 knots, it is very mild a and slow.
The R44 is much more stable and the right turn tendency is much less at high speed.
I hope this helps.