In controlled training conditions, I have yet to see anyone lose less than 10% Nr when counted down into a practice double engine failure. I would suggest that the situation would be impossible to recover from were it to happen for real.
There was a thread running a month or more ago in which Nick Lappos (I believe) very capably explained the whys & wherefores of designing twin engined helicopter rotor systems that are maximised for "powered" operations, rather than single engined machines that tend to make more compromises towards power-off flight. I would not chose a twin engined helicopter based on its double engine failure characteristics, nor do I fly with thought of such infrequent failures on a daily basis - yes, of course the mind becomes more focussed in the event of any emergency or system failure.
I would concur, however, with the switch comments - how, especially, it was decided to put the Cat A switch so far from single pilot reach, I cannot fathom. (For those unaware, this requires hand off collective at 55kts / 100' or so, on take off or approach, unlock seat harness, reach forward, make switch, re-sit & continue). Why one of the unused buttons, on either collective or cyclic, can't be used is anyone's guess!
There may be plenty of things to improve on the 135, but as JTT says, better be able to use working bits, in a fiddly fashion, than just fiddle with a few bust bits in a hangar queen!