It's something I know I am interested in. For me, I don't really care if my statement doesn't cancel until the very end if there is a RI violation. The benefit of not having deletes be slow on tables which have others referencing it with a fkey which don't have their own index is huge IMO. I have a good number of those type of logging tables where an index is not useful 99% of the time, but every once and a while a bulk delete needs to happen.
It is far from a premature optimization IMO, it is super useful and something I was hoping would happen ever since I heard about transition tables being worked on.
note: my sentence about premature optimization was related to my idea to divide RI check per 10K rows.
