Thanks for the cool suggestion. That is a pretty awesome idea. However, this may not be too feasible since badger (the KV store that dgraph uses) is pretty tied to file system semantics, and might not work too well in a serverless environment, but never say never :-).
In the meanwhile, i don’t know if you saw my response to your other thread. We are coming out with a pretty cool solution for a managed GraphQL solution, which should be able to act as a primary data store for your cloudflare worker.
Also, as a shameless plug, if you are looking for a great an easy way to get started with CF Workers, you might find this project of mine helpful: GitHub - gja/create-cloudflare-worker: A starter kit for building a Cloudflare Worker