Oh dear, my head's hurting now (not hard)
Skycop quite rightly says that if a single engine french a/c lose power it will yaw to the right. If a twin engine one lose one engine then you will have a temporary power setting somewhere between that required and zero. However you will still, unless blessed with superhuman reactions, have a pedal position suited to the original power setting ie more right pedal (French still) than is needed for this temporary new power setting. Ergo the a/c will yaw to the right. Now the length of time it takes for the power to be restored will naturally depend on the sophistication of the engines ie Arial with FADEC will respond v. rapidly, PT6 with hydro pneumatic AFCU will take about 4 seconds.
The a/c in question undergoes a steady and pronounced yaw to the left through about 90 degrees before ditching. For this to be due to loss of Nr implies gross overpitching to the extent that an Nr of perhaps 60% was reached. To suggest that sufficient power/Nr was then available to lift the a/c clean out of the water (extra drag doing that as well) is not plausible. A brief lurch maybe, not a fully fledged break dancing display.
Old beefer, your name and location make it obvious you have acces to the odd 412 or two (sorry, Gryphon), have you tried looking at TR response at low Nr? I've had demonstrated to me an OEI take off in the 212 using 'blips' to 70% with no problems. That's 49% of normal TR thrust. I've also tried hovering manouevres myself OEI down to 75% with no ill effects. I still find it hard to believe that the onset of yaw witnessed in the a/c in Taiwan could be solely due to Nr problems. I would support the idea that the subsequent pirroueting (spelling) was caused by post impact TR drive damage.
------------------
Another day in paradise