JimL - you make it sound like an autorotation is a. dangerous and b. difficult - even at MAUM it is a straight-forward manoeuvre, especially straight in to a runway.
Consider your proposition - one engine has failed (for an undetermined reason) and you elect to fly a normal approach to running landing on your runway. During the 60 seconds on finals (I think it would probably be a bit longer), the second engine decides to let go as well and you crash into the undershoot trashing the aircraft and injuring/killing the pax. What exactly is your defence? I was following the RFM and anyway we don't practice autos because it's not good for the airframe?
I'm not saying this is the only way of doing it or that it would be possible or practical in all situations but I offer it up as an airmanship consideration to those who might not have considered this course of action because they have only been conditioned by their civilian experience.
I am clearly not alone in my thinking vis Bertie and Brian so maybe those who were so quick to condemn out of hand should maybe pause for thought.
Roundwego - in the many hours of single-engine helicopter flying I have done, the consequences of an engine fail were seldom far from my mind but as long as you kept out of the avoid curve (H-V curve) you could operate safe in the knowledge that you could enter auto and make an EoL.
A heavy twin full of pax on a single engine would have a much bigger avoid curve and so the procedure for flying a SE approach should be modified to accommodate that.