Nik: This is an outstanding guide. Arguably the most comprehensive I've seen (and one helluva lot better than the laughably-worthless offering by MDN).
I've personally used IndexedDB for several years, now, for tasks ranging from local storage for the client segment of an MPC encryption identity token to a 1:1 mirror of a full Jira implementation (to permit more granular queries against the dataset). Even wrote a full SQL-syntax abstraction layer over it, once, and am currently in the midst of writing a dynamic-hierarchy proxied abstraction layer over it as we speak.
All that said, I STILL learned quite a bit from this. I had no idea some of this was the case (and, while it has necessitated re-writing a swath of my library code, this is gonna prove invaluable for some of the querying algorithms). THANKS!
Sub'd and followed!