The whole idea is that one parameter remains fixed - aiming point. The other is continually corrected.
If an approach angle is flown to an aiming point which allows the required constant approach speed to be maintained with a power setting somewhere between idle and max, then the technique can be used. It worked quite happily in the Bulldog when we were teaching 'short landing' technique with the IAS very close to the 'wrong side of the drag curve' - it just neede rather more prompt power application and more frequent IAS cross-checking.