It feels like a firmware problem to me, too. I'm expecting it to be fixed in a future release, and, of the bugs the software has, I hope it's the one with the highest priority.
It feels as if something is spinning, waiting for data that isn't going to come, because the other end thinks it's done. I base this only on the observation that when it gets hot, opening and closing the thing (which obviously triggers some comms between the CoDi and the main device) can fix the problem. And it usually starts again when something has triggered something which causes something to display on the CoDi.
Might be totally wrong, of course. Without the source it's impossible to say.