Jim Bosch what do you think we should do with this ticket?

I don't really recall what the intent was well enough to attempt the ticket myself, and I'm also dubious of the value of writing new hypothetical use cases for middleware when we have plenty of very concrete ones breathing down our necks.  I'm inclined to Won't Fix it.

I would say that it's a very good idea to be able to roll back a schema upgrade if something goes wrong. That could either be via "round-tripping" (upgrade and downgrade scripts that manage to not lose information), or it could be via backup/restore, as long as the problem is identified before any new data has been stored in the schema.