Just adding my vote for prioritizing development for multi-graph/database access.
Consider a distributed database model with graph subscription instead of read/write access to a singular graph. Imagine graph access linked through on-demand subscription, allowing multiuser concurrent access. Enabling graph sharing while compartmentalizing information with clear delineation of relevancy and access control.
Person X Personal Graph (Generally Private)
Person X Company Graph (Generally Relevant to Person X Work)
Company X Graph (Company or Department Relevancy)
Project X Graph (Relevant for Company X Project X)
Person X Family (Family X Shared Graph)
Querying information would search all subscribed graphs or better yet some sort of database linking, while writing could allow duplication of information across all selected graphs for redundancy or blurred lines of relevancy.