Torsten,
Thanks for the insightful comments. I will try to clarify a couple of the points you raise.
1. I'm not sure if you regard matter-energy to be auxiliary to spacetime, or if you regard the two to be part of a single fundamental structure. I far prefer to regard them as part of a single structure, which I describe at the classical level by a binary relation. Hence, I do not expect this structure to be manifold-like at sufficiently small scales. This forces the theory (if it ever becomes sufficiently developed to be called a theory) to predict 4-dimensionality (and many other things) at large scales, probably by means of action principles and entropy.
2. The causal set theorists have done a lot of experimenting over the years with "sprinkling" points in 4-manifolds; as you point out, the binary relation doesn't determine the geometry, but the argument Rafael Sorkin makes with his "order plus number equals geometry" phrase is that you can recover 4-D geometry from a suitable order if you supply appropriate measure-theoretic information as well. I think this is true. If it is not, then my ideas probably don't contain enough information. Note that the causal set theorists make a lot of other assumptions I find dubious, however.
3. Regarding fundamental theories and single entities: the desire to describe spacetime and matter-energy as part of the same structure is a lot of the motivation for my ideas. I call the classical "posets" (not really posets in general, of course) "universes" to emphasize background independence: in Feynman's sum over histories, one thinks of particle "trajectories" but generally ignores the obvious fact that the "underlying spacetime" actually ought to respond in different ways to different trajectories, so one is summing over entire "universes" in this sense, not over trajectories in a single "universe." However, the "Universe," which is quantum mechanical, is the entire family of posets with their induced order. After all, similar remarks could be made about manifolds; the etymology even reflects this. A manifold is a set with an atlas, but no one argues that the presence of multiple charts means that the manifold is not a single entity. This analogy is imperfect in multiple obvious ways, but the main point is just that different models partition information in different ways and it is not necessarily easy to uniquely define what "unified" means.
4. Regarding your final point about the open future, a single classical universe contains its entire history, but such a universe may be regarded as the source of any number of different transitions. In this sense the future is open and the past is fixed. However, I suspect this may not entirely answer your question.
Thanks again for the feedback,
Ben