Tail Rotor problems
First of all there is no such thing as practising tail rotor failures in the real aircraft. You can simulate pedal seize and have a half-hearted attempt at drive shaft failure but you would need to be well insured, very brave (or foolish) and have access to a time-lifed airframe to go off and 'de-clutch' the tail drive shaft.
I am blessed in my job to have access to a very expensive level D simulator that is programmed with the best mathematical data available to simulate a drive shaft failure and pedal seize at any phase of flight - but it is just that, a mathematical projection.
I caution my students to be careful what lessons they choose to take away with them after sim sessions where tail rotor failures are featured. A survey carried out by the UK CAA about 10 years ago revealed that an analysis of hundreds of worldwide TR failures only 30% were drive shaft failures. The rest were a diverse mixture of tail rotor component failures. Putting too much emphasis on drive shaft failures can lead you astray.
If you are going to develop a single instinct as a result of training in the my sim I hope it will be to lower the collective to the bottom stop as soon as you identify loss of directional control - no matter what phase of flight. Subsequent careful re-application of collective may be possible depending on airspeed. Delay in lowering the collective will result in an increasingly dynamic departure from controllable flight. The events in Macau would appear to reinforce this view and as SAS points out the very unpredictable consequences of the a TR blade departing may result in dramatic effect on CG. Maybe TRIs in other types/sims would comment (agree or disagree) on my suggestions.