In the cruise, day VMC, at a suitable height, just about every emergency is benign and can be dealt with in the relaxed manner suggested by some posters. But at low level, night, in the hover over the sea and many other health risking conditions you do not have the luxury of calmly assessing and diagnosing the problem - you must be trained to react (not blindly but quickly and correctly).
All the hours spent diagnosing runaway ups/downs/unstable governing during training (either in the sim or the aircraft) are then focussed into a motor programme that will hopefully save your life.
The same goes for engine failures on singles - easy enough at 1000' on a nice day but at 200' on finals to a rooftop site?
Hyd failures in a squirrel - relative piece of ease over flat ground but in a confined area?
I understand the attitude 'it hasn't happened to me so why did I bother to train for it' because accidents always happen to other people- right?
Some helo systems have their own little gotchas - Lynx engine oil pressure for instance; spend more than a couple of seconds looking at the pressure gauge that has just dropped into the amber and you will have a catastrophic seizure of the reduction gearbox to deal with.
Know your aircraft systems and train as hard as you can to deal with emergencies, think through worst case scenarios and how you might deal with them. Try to get feedback from other operators of your helo type about failures they have had which might be waiting round the corner for you!
Alternatively hope it doesn't happen to you and hide your emergency checklist at the bottom of your nav bag - just don't pretend to be a professional aviator!
If a low cost sim was available for the light single market it would probably save a lot of lives and bent aircraft but simulators only come into their own when the relavent aircraft is expensive to operate or has many complex systems (clearly not the R22 then).
The only thing to beware of with a simulator is the 'it happens in the sim so it must happen that way in the real aircraft' mentality - a simulator can only react in the way it has been programmed and if the real data concerning rates of pitch roll and yaw has not been gathered from the real aircraft during the real emergency (TR failures for example) the the sims performance is only a 'best guess'.