Of course it matters. These theories prescribe different remedies. If the model requires assuming that some law applies, as you suggest, or that "incrementation" is functionally additive, as EY implies, these postulates should be explicit in the model, as they're crucial to deriving the remedy.

I don't read postings like the original as asserting models, although lukeprog might disagree, and to the extent that he would, that's the sort of thing I generally just tune out. I regard them as being more like things such as MBTI or astrological types -- schemas for imputing structure to some Rohrschach blot of a phenomenon as a method of generating ideas about it. And it doesn't matter where such ideas come from (which is what I had in mind when saying "it doesn't matter") if one gets practical use out of them.