Each painter's isomorphic image in Z mod 2^17 - 1 is that painter's effect on zero. Profound idea, really. Just lacking a proof...
Perhaps a good start would be to show that the painter permutations are, themselves, isomorphisms; ie, show A(x + y) = A(x) + A(y).
Nope, scratch that; can't be, because A(0) is never 0. Back to the drawing board.
So, the basic question is this: I have 17 functions, commutative under composition. What do I need to show that those 17 functions are elements of a LARGER group isomorphic to Z mod 2^17 - 1?
And why is x_0 = 0 = 000...000 so important?
I think I'm starting to see it. Let f be the "tenative" isomorphism from the "expanded"*** painter permutations to Z mod 2^17 - 1. I'm saying that for all A in {PP}, f(A) = A(0). I think it suffices to show the following two statements:
1. Whenever A(0) = x and B(0) = y, for any A, B in {PP} and x, y in Z mod 2^17 - 1, then AB(0) = BA(0) = x + y. And its easy to show the commutativity part.
2. If D & E are in the "expanded" {PP}, and D(0) = E(0), then D(x) = E(x) for all x in Z mod 2^17 - 1.
***And by expanded painter permutations, I mean all 17 painter permutations, PLUS all their compositions. In other words, the group generated by the painter permutations.