Remember guys, the ECL's would not have to come all the way back to IDLE. If they even came out of the "fully open" position it might cause the torques to go to zero and the NR to wind down . Unlike our fixed-wing brethren, helicopter throttles only have one working position: FULL OPEN. Anything in between can be considered OFF (or IDLE but in any case no power is delivered to the MR to sustain flight). Is the C++ any different?
Hypothetically: So the windshield post fails and the windscreens fail inward, knocking the ECL's back "just out" of the fully-open position. Off go the torques, down goes the NR, but the crew is preoccupied with the windshield failure and doesn't get any immediate notice of LOW RPM or ENG FAIL (because the N1's are still above 66%) until it's too late.
We can only imagine how confusing, chaotic and disorienting it must have been in that cockpit from the time of the "loud noise" (whatever it was) to the crash.
By the way, if memory serves, the S-76 crew was not restricted to 700' outbound from the base. Unless I'm mistaken, the general rule is "500' inbound, 700' or higher outbound."
Helicopter pilots will give you a million reasons they like to fly low...why it's even necessary.