Sounds like a very clever idea- provided he left enough notes that a reasonably experienced engineer can pick it up and carry on. Its all going to depend on just how much code is left over when the mutually exclusive options are turned on in the right way - obviously some of the code is most definitely not used when other sections are enabled.

While there is no reason why you cannot separate off the interrupt code, it probably makes more sense, when the interrupts might be intercepted by the OS first, to put their code in the OS section.

I haven't heard anyone trying to use OOPL in assembler. The baggage and bloat is too great to do anything useful.

Steve