Zimmer,
What Nick refers to is known as "off model" responses. This is what happens if the simulator programme does not include sufficient (or incorrect) data to accurately represent what happens for real and it is a well-known issue.
A simulator is, after all, a big boys sophisticated computer game built for a particular purpose. The sim "aircraft" does not actually have the properties of a real aircraft unless those properties are accurately included in the programming. I can assure you that the simulator company does not go out in a real aircraft and fail the tail rotor in order to gain data (obvious really, innit).
Therefore, what is programmed in is a representation of what is assumed to happen. Yaw rates and other effects following on from those rates may have to be mathematically derived and are a "best guess".
A few years ago, I worked alongside a simulator software guru who worked on site at a simulator we were using, as a major customer. He had written the programme into the simulator software. We had some particular problems because it was a new setup and we were using the sim for critical failures that no one had tried much before. The simulator response was definitely different to what the real aircraft did. On a number of occasions I explained the differences and stood next to him as he plugged in his keyboard, typed in raw machine code to change the programme and he had modified the response of the aircraft just like that.
Just bear in mind that a simulator is great for procedures and working out techniques but it may not always be completely representative of the actual aircraft in its response. For example, the helicopter simulator I instructed on would fly a full fixed-wing aerobatics sequence, including inverted flight and slow and four-point rolls! I used to demo it for fun and a few pilots came out both shaken and stirred. I always warned everyone that under NO circumstances should the manoeuvres be tried in the real aircraft because they probably wouldn't have got through the first manoeuvre in one piece.
However, simulators are still the best medium we have for training for unusual emergencies but these possible limitations of the system must always be borne in mind.
A "flying by numbers" approach to any particular scenario in a particular simulator may or may not be 100% valid. I have seen that using two different manufacturer's simulators of the same aircraft type may give slightly different ideas of how best to cope with any one scenario in detail!
ShyT