Ballscrews can in general be back-driven, the difficulty depending on the lead.
Good practice would include fabric gaiters over the exposed portions of the screws to protect them from external contamination, and to limit the spread of internally generated contamination. Gaiters don't weigh a lot. Ordinary ballscrews do. Aircraft may use hollow and/or nonferrous ballscrews, which are not otherwise common.
I recall examining a reactor control rod drive (some decades ago; they may be different now) comprising a roller screw, the nut of which was split on an axial plane, and spring loaded and hinged so the rollers would be pushed out of engagement on power failure, allowing the control rod to drop, scramming the reactor. I think the magnetic field of the drive's motor pulled the nut halves together, so the nut and screw were engaged so long as the motor was powered up. It was relatively simple and sturdy, and definitely not flight-weight, and definitely not affordable by ordinary mortals. I'm not aware of anything similar produced for commercial applications, but it may provide food for thought.
Speaking of which, it should be possible to build a clutched linkage, e.g. using trailer brake (electro)magnets run at reduced power to grip a flat ferrous surface (you need a pulse of higher power to pull them in at startup), interposed between some kind of actuator and the linkage to the flap itself. Startup would also have to include some procedure to align the parts, e.g. by letting the flap drop against a stop (or be driven to a known position by weak springs) and driving the actuator to the corresponding position before engaging the clutch.
Mike Halloran
Pembroke Pines, FL, USA