I agree with Huajie and would also add that the Two-Stage Procedure (Section 184.108.40.206) does not explicitly address a change in Importance therefore I would assume this is not anticipated – or maybe that it was discussed and the authors could not envision such a scenario. In adjusting the forces per the ratio R/ρ for the upper and lower systems (Section 220.127.116.11, item d), note that Ie is not included. Would you augment this to also include the ratio of Ie for these systems? The ratio of these two (R/Ie) is found pretty much everywhere else so being that it is not in this section allows for an interpretation that the authors thought that a change in Ie couldn't exist when using this procedure.
I mention this to add to Huajie's point that collapse of the upper portion could easily be argued as not 'independent' to the lower portion (i.e., 'independent' being a key word used in the Provision). Unless this sort of thing is precluded, my opinion would be to use the larger Ie for all. Even more directly to your question, yes this is allowed horizontally but no I don't see this working vertically unless provisions are made to assure the systems are independent (i.e., with respect to failure).
The building official may allow it but before inquiring you might consider if you're capturing the intent expressed in the Commentary 1.5.1 (as you've probably seen there is no commentary for 1.5.2):
A rational basis should be used to determine the Risk Category for structural design, which is primarily based on the number of persons whose lives would be endangered or whose welfare would be affected in the event of failure.
I'd expect that this will help you decide if you even want to pursue it. If so, I'd think you'll need to have a good answer to this for discussion with the building official anyway.
Hope this helps,Keith