Methinks the FADEC team writing the software to protect the donks never had the Hudson in mind - up close and personal.
You are correct and by FAR regulation they have to do that, else it couldn't get certified. In the old steam driven days with knobs that a pilot could twidle the design regualtions did not regulate pilots so the pilots felt empowered.
Well in todays glass cockpits and computer driven era the pilots are still empowered but alas they don't have the same knobs to twiddle anymore and they pretty much have to sit and watch the thing do whatever it's gonna do.
The safety related presumption is that the design regulations provide sufficient protection (within a 1 in 10,000,000 probability) and by taking the pilot out of some controll you gain by reducing pilot error.
Yes the controll logic needs to be examined and reported (my earlier question: what was it controlling on see post # 1179 )
In the end the issue comes down to the bird design certification standard and the balance between capability and need to avoid. By regulation you can't design a FADEC to take the engine outside its certified envelop