Changes

Advantages: users could choose a small/simple server or a complex one, irrespective of the front end (and so save RAM, or gain speed, or features). Developers could benefit from a pair of smaller projects, not having to worry about complications internal to the other section; and making the change could nudge certain long-standing problems to be thought out to make a neatly organized plan, even if the implementation of them in each end still needs work.

I think (maybe I'm wrong) that this is an easy transition plan for what amounts to significant re-thinks in message storage, and I think it isn't far from some things already in motion, but I don't think that direction is enough of a priority at the moment.