Omnocracy v. Semantic Web

I've been reading through some of the documentation for RDF and XML over the last few days and have come to some conclusions about the similarities and differences between the "semantic-web" concept as it currently stands and my present vision for a reputation based annotation community.

1) both recognize the need for the atomization of data (both internal and external to the system), the use of URI's for identification of pieces and the need for a subject/predicate/object syntax to foster use of these resources.

2) both recognize that this is all useful to facilitate a broader sharing and remixing of information.

3) both recognize (I believe) that at some levcl, schema's need to be defined as separate layers so that "data-structure" is not hard-wired into the data. Without such separation, remixability and re-purposing of data would be severely inhibited.

4) However, I believe the semantic web is a different layer in the hierarchy than the omnocracy system I am attempting to design.

5) Omnocracy requires the structural attributes of the semantic web, but is designed to do something different. The decentralized nature of the semantic web would likely inhibit the processing time for a layered processing of data from n-nodes on n-layers. As such, I do think that some sort of centralized storage system may be required for Omnocracy to rapidly return results. It is most certainly possible to implement omnocracy in a distributed form, but this would almost certainly be accompanied by large time delays when complex processing occurs. Perhaps, there could be a centralized caching of all material (google-style) for rapid processing for information that required it. At the same time, this increases infrastructure costs (for the central firm housing all this) tremendously (i.e. there would be virtually none otherwise).