If you really start out with nothing, then there is no preexisting information. You must first tell cancellation how to occur, and tell zero how to "fission" into two or more mutually cancelling entities. Since these "instructions" on how things should cancel consist of constraint - of restrictions that limit the nature of "nothing" - constraint has logical priority over cancellation. So we have to explain how the constraint got there in the first place, and we have to do it without appealing to any preexisting cancellation operation supposedly inherent in "nothing".
By default, the entities comprising the system that you call "the sum of components within the space-time manifold" must themselves combine to form the constraint, and since neither the constraint nor its constituents equal "nothing", we're forced to introduce "something" at the very outset. This tells us that we're not really starting out with "nothing", but with unbound potential or UBT... something which, by its nature, contains everything that can possibly exist. This amounts to interpreting 0 as "0 information" and replacing it with a combination of unity and infinity.
I.e., by interpreting 0 as the absence of something called "information", which implies that it equates to superficially unrestricted homogeneous (and therefore unitary) informational potential, which in turn equates to existential infinity, we're making it possible for 0 to represent the sort of cancellation point you envision. Since this implies that existence is a self-defined attribute - after all, there is no preexisting informational framework capable of defining it - only possibilities with the intrinsic ability to define and sustain their own existences can be actualized. This is what it takes to get "something" from "nothing". Source: http://web.archive.org/web/20020619055425/http://www.ctmu.org/CTMU/Discussions/CTMUForumZero.html