Uh, Zelos, I don't understand what it is you're arguing anymore. Is it that momentum is always conserved in non-curved space time? Then I think I agree with you, although I don't agree that that actually tells us anything useful, since as PY says, friction and other things have to be measured on the macroscopic level. Momentum is really only useful directly before and after a collision.
If you're arguing something else, shame on you!