Column-oriented has better performance for analysis but worse for highly transactional workloads. I think standard operating procedure is to dump the whole database every night into a column-oriented database and then do their analysis on the second database.
The comment which I deleted contained something along the lines of "why does this work and why would a user have to concern themselves with this; a table flip seems feels vaguely like something that should happen as automatically as compiler optimisations".
This response does not really answer the question that is still just half baked in my brain, but since I did get a response (cheers!) restoring the record seems like the right thing to do.
