MartinIsti
Databricks Partner

Couldn't agree more, TamD. My impression is that DLT (that has strong ties with LakeFlow Connect) is being heavily developed right now and the documentation is not necessarily keeping up with the actual code-development. It should and it's not a good enough excuse if this is indeed the reason but that's realistic.

I would think with new features many current forms behaviour can easily change and this storage might be one of them. Especially when you think of DPM (Direct Publishing Mode) of DLT where you can dynamically specify target catalogs and schemas.