Geoffers - I think the SL is a bit of an a*se-covering exercise as it highlights that correct pilot action (ie full right pedal) will stop the spin. This was ably demonstrated by a French TP to put to bed the myth of Fenestron stall on the Gazelle - he established a 120 deg/sec rate of yaw and stopped it dead with full right pedal (and a little bit of an overtorque, but that is better than crashing).
Note the SL refers to ALL TRs, not just the Fenestron and simply shows the thrust response from pedal position is essentially linear with a conventional TR but more S - shaped with a Fenestron.
The SL primarily concerns itself with occurences in the low hover where incorrect pilot action could easily result in trying to land with a high rate of rotation - not usually a good outcome from that - the Taiwanese 365 was in a high hover so LTE/FS is very unlikely to be the problem since it would be easy to flyaway from it.
However - Sid's link (which I can't get to) might be a possible answer - a big slug of seawater through the fenstron could easily trash the TR driveshaft due to the sudden drag.