This is a very intresting subject and I have to send my 2 cents again.
BB I think the problem you refer to was from an accident on Greenland in the mid or late 80 ies. This was due to (if I remember right) a broken drive belt to the hydraulic pump in very cold conditions and with a new style of Spherical Bearing (load carrying rubber bearing) fitted. Not to the fact of slipping drive belt.
The fix was a new style of bearing, and a new lower temp limit of the previous one.
The RFM also calls out for warming up the spherical bearings before takeoff in cold conditions by moving the cyclic 3-4 cm fwd for 2 min and to check the force to move the cyclic without servo.
We have been flying B, B1, B2, B3 in very cold conditions (down to lower limit) for a long time without problems, we started with B in 1979. We even tested a cog belt driven pump for EC, but it never went into production. Even if the belt drive looks very simple, it has just failed on us once in all these years, and it was a successful landing afterwards.
I have always as a training captain had the opinion on single hyd system helicopters that if there is a procedure in the emergency checklist that calls for turning off the hyd system ASAP, then it should also be trained at cruise speed. Normal procedure is to bring down the speed to around 60 KIAS and then turn it off (even at ECF they do not train this at high speed if they have not changed there way of training). But if you get some sort of hardover you need to get it off immidiately, no matter the airspeed, without having tried it before it could be intresting
Therefore when I train someone on the 350 I demonstrate this and then let the pilot try it with increasing airspeed up to cruise. At cruise it is hard work, but since the cyclic want to go right and aft, it is very much controllable. But since many 350 pilots from time to time hold the cyclic "very light" to get good feel, I think it is important to know what they are up against if you turn off the servo at cruise speed.
CB