Document schema dependency around datasets and projects

[dosumis]

The current API code assumes each dataset has a project and the projects link to person nodes (with ORCIDs) to encode permissions. The aim of this ticket is to document that schema so that we can

(a) Manually extend KB.ids for initial tests
(b) Add wrapper code to VFB_neo4j for managing projects & people in future.

[dosumis]

TBD: Do we need projects? Can this be better managed via direct links from datasets to people. Initially this will be a link between dataset and it's creator, but we might allow the creator to share access in future.

If this is all user managed, data will initially not be published. User should have the ability to choose to publish, at which point a review by VFB will be triggered.