The rudder inputs appeared to be over reactive and late : I thought that the pilot needed a basic refresher in er basics.
Seems to show that auto is auto and is apparently incapable of predictive application.
Perhaps an example of the challenges facing fully auto systems
Qué?
This is a manual landing. The automatic rudder inputs happen airborne, and are the FBW system's response to the varying crosswind, which it detects as sideslip. It doesn't care about runway alignment, or anything else....it just wants to get rid of sideslip. As a side effect, these constant corrections tend to remove about half of the effect of the crosswind changes.
There are very few lateral corrections during the video, so the aircraft has maintained the centreline reasonably well, without having to chase it.
The landing itself is hard. There is no flare that I can discern. It is also early (at about the '500 foot marker), which also fits in with 'no flare'.
The rudder inputs on the ground are manual. Recall too, that once on the ground the rudder pedal inputs are also translated to nose gear steering. Basically what you are seeing is over controlling. Landing with the drift intact is not an Airbus procedure. The 5º drift limit corresponds to about 12 knots of crosswind. Boeings are allowed to be landed with drift intact right up to their crosswind limits (which will give about 15º of drift). The initial aircraft energy vector is down the runway, and as long as you don't let the nose gear come down, the aircraft will skid in that direction, giving time to smoothly use the rudder to push the aircraft into alignment.
The aircraft is huge, and does not respond quickly to anything. It has a lot of inertia if you let it go astray, and it can be very hard to catch.
Your comments re automation have no relevance to this video.