IIRC, Group III additive response is almost as bad as PAO, it also has horrible solvency but it doesn't have the seal shrink tendency PAO has. They both, because they are ultra-pure, have the same issues except for that last one, which requires more "accommodation" via formulation as you noted, using AN's, Esters, or both. The Mobil blending guide examples typically used a small percentage of ester for this purpose.
Of course being "not as bad" has its advantages and makes Group III easier, or perhaps "less challenging" might be better phrasing, to blend with. PAO, by itself, also has better oxidation resistance, but that can be dealt with via additives for Group III nowadays.
PAO's main feature, and the reason HPL (and others) use it in their 0W-xx's is of course that it has no wax, so there's nothing to crystallize like with Group III, that requires PPD's. This is why the pour points are so low and this unrivalled cold temperature performance and high VI offsets the low VI and relatively poor cold temp performance of the AN's.
One thing we discussed in the past (
@Shannow myself and some others) were the requirements for the Winter rating, which allow an oil to "slip" a Winter grade in service. PPD's, like VII's and other additives, do degrade in use, so if your cold temperature performance is predecated on the function of PPD's, it would not be unusual, and would be acceptable in fact, for that performance to degrade to the point that it no longer met its original Winter grade in use. So your 0W-20 could end up a 5W-20 for example, or your 0W-40 a 5W-40 (or 5W-30 if you have VII shear too).