Sorry. PBL, I have to support XPMorton here. The yaw-roll-decoupling cannot take ' a holiday', since there is no 'real' yaw-roll decoupling as far as I have understood Airbus FBW control law logic! But well, correct me if I'm wrong....
If a roll is commanded, Normal Law will adjust pitch to provide additional lift to keep level flying as well as rudder for a coordinated turn.
But vice versa, if you apply rudder, the A320 will (initially) exhibit a roll due to yaw-roll-coupling as any other swept wing aircraft. The control law does not automatically and simultaneously compensate this roll moment by opposite aileron deflection.
But yes, the FBW control logic will nevertheless react to this roll, but different as you might assume reffering to a true 'yaw-roll-decoupling'.
As far as I understand, the Normal Law provides 'roll stabilization' as long as there is no stick input, i.e. it will counteract any uncommanded roll and try to return to previous bank angle. Thus the Normal Law does 'see' the rolling of the A/C due to rudder deflection like an 'external' disturbance resulting in an uncommanded roll ('as long as the stick is untouched).
So there is a kind of 'yaw-roll decoupling' through the back door.
But this is a significant difference to a true 'feed-forward' yaw-roll deccoupling where the controller would know in advance about the A/C behaviour to a rudder deflection and thus would apply the appropriate aileron. But if the controller sees the roll as an external disturbance, the controller does not 'know' about the size of the disturbance and thus not the required control action. The reaction will be a result of actual/nominal value difference, applied control law, and gain factors and thus can be significantly slower up to the fact that depending on the implemented law a finite residual error might remain....
And thus as the manual cited by XPMorton points out or Chris practical experiences report, into the wind aileron has to be applied to keep wings level or the upwind wing down during fast decrab.
And I guess we don't need to discuss that rudder on swept wing A/C is a very strong roll control that can even outperform aileron capacity (remember 737). So no wondering that the wing coms up fast durind this 'significant' decrabbing'....
Regards, ihg
P.S.: I don't believe in the gust story....mean cross wind component was high enough to have the A/C drifting across the runway as seen ...