I really see no basis for an approval on a not fully-substitutable basis. The FDA by doing this would be saying "it's sort-of the same" and if they do that they have no real basis for claiming equivalent safety and efficacy. There is no comparable situation where they have ever done this before - I believe the existing non-AB generics were all designed to be different than the originals in some way (dosage form, PK, etc.).
You may very well be right. I'll let Dew or others weigh in on whether or not there could be any type of comparable situation as I am unaware.
To my mind it's either a full approval in the near term or a significant delay (longer than a year) before approval.