BOAC, it's the subtle differences that when not thoroughly understood make the debate heated sometimes. Unlike 737 Classic (never flown NG) AB has pilot selectable TRACK function that can be used in lieu of normal HDG lateral mode.
By definition the TRACK reference mode will adjust for drift in changing wind conditions - within the technical limits of a man-made design. This actually answers the original question.
Coupled to TRACK mode is Flight Path Angle mode that does pretty much the same thing in vertical plane. So you have a choice to use either HDG-V/S or TRK-FPA. In TRK-FPA reference a small trajectory symbol is superimposed over PFD called flight-path-vector, commonly referred to as "bird".
Anybody could fly ILS beam using TRK-FPA (or HDG-V/S) if there were a reason to do it except of that being a grossly silly idea. OTOH, TRK-FPA (or bird) is designed for NPAs - for NDB APCH once on final track you select TRK to appropriate figure, at FAF FPA do required value (i.e -3,3 deg) and irrespective of wind and GS changes everything pretty much takes care of it self.
The subtle difference from 737 (suppose) is the fact that AB designers decided not to introduce VOR-beam (radio based) tracking function to FD. Instead of VOR/LOC pushbutton AB only has a LOC pb. So to track a VOR radial the procedure is to activate "bird" and once established inbound set INB CRS and in theory no further adjustments are required.
In real life small adjustments within 1-2 deg are required to keep the VOR raw data centred. Possible reasons:
- radial is scalloping so changes needed to follow the imperfection
- radial is orthodromic (curved track) by definition whereas TRK flies straight ahead
- once displaced (by gust that was not perfectly compensated by the guidance and AP due to AC inertia) from the ideal centre, TRK would continue parallel to radial unless corrected
- momentary drift of the commanding IRS
The IRSs drift exactly the same as on Boeing, probably owing to the fact that they are identical piece of equipment. Residual GS when stationary on ground is typically 1-3 kt. If such drift is 90 deg to desired track to be flown it indeed needs to be compensated conventionally. Of course this doesn't affect the NAV accuracy / FMS position as that is derived from MIX 3 IRS data that is constantly updated by DME/DME and then pinpointed through dual GPS with RAIM. Again, I would be surprised if Boeing system were any different.
Hope this removes some of your fears
Yours,
FD (the un-real)