BOAC, OK465, RRR, A33Zab;
I believe the reason for ensuring the disengagment of the FDs during the UAS drill is based upon the default longitudinal AP/FD engagement mode of V/S or FPA. The removal and reappearance of the FDs would logically be the same as pushing the FD switch OFF then back ON again. When that occurs the V/S or FPA is synchronized with the aircraft parameter, whatever it happens to be at the time of re-engagement. From there, the pilot makes an appropriate selection such as OPEN CLB/DESC, or, for immediate altitude hold, setting the V/S to zero, etc. (Another way to put it...V/S or FPA are simple, selected (ie, not managed), modes - the FDs could not "know" what "1400fpm" or "6000fpm" was for...it just reflected current conditions, from which selections may or may not be made).
The essential aspect of the UAS drill is flight path stabilization via pitch & power. The FDs could non-intuitively counter that response depending upon conditions at re-engagement.
Mountain Bear;
Complex systems fail in complex ways but that's no excuse for making hash.
Ensuring a stable flight path would have avoided the accident.
While the slavishly-following-the-FD theory makes sense, as posited by some even before the report was issued, it makes no sense to many who have flown bread-and-butter transports that could actually be stalled any time and who now fly this or any aircraft with complex automation; it's a tool in the kit and nothing more - still is. "Looking through" the directors to the raw data on the instruments and combining that with one's understanding of the energy and stability of the machine was always a valued approach but not anymore.
The ability to understand and even have some depth of intuition about flight in heavy transports in the operational environment is the larger story here.