Thanks for looking into this, Terry.
Regarding the rules. I deliberately created them that way and order because of the "first match wins" logic, like e.g. my pfSense works, or my switches handle ACLs. That's the reason why e.g. the 1080p rule, individually, doesn't make much sense. Only in the context of all rules, and the order in which they are organized, does it make sense and works. Until now I had no reason to doubt that ZoneSwitch honors that logic. All files always end up in the correct zone. Based on your edit, I think you also consider the rules correct (in their entirety and order of course), right?
It feels like sometimes DisplayView doesn't realize which zone starts with the playback stays with another zone (thinking about it, maybe I should've picked another title for the thread). That brings up a question. What part of MC is the authority for the "current" flag of a zone (assuming that "current" equals my "active" designation)? Maybe, whatever makes a zone the "current" one is having an issue with something.
It might be related to *.iso files because, from what I recall, it happened always with them ("always" meaning that in the few cases I observed, it was "always" an *.iso file that was playing *). But. Several "buts". First, I cannot reproduce it reliably yet (trying again later). Second, in all these cases, the playback had been initiated in the correct zone, but DisplayView doesn't switch over to that zone automatically (so maybe it's a DisplayView issue?). Third (and please correct me if I'm wrong), doesn't MC "see" *.iso files as real discs, once they're mounted? At least they do show up in TheaterView and the zones itself (StandardView) as a DVD or BD movie, respectively.
(* Also, in my case, the video part of my library consists of ca. 99% *.iso files, and then a few *.mkv and *.mp4. So, the probability for encountering a random issue while using an *.iso file is pretty high anyway ;-) That's why I'm a little hesitant attributing it to *.iso file handling, before having a way to reproduce it.)
To cut a long story short: I need to find a way to reliably reproduce it. Before that, any testing especially on e.g. your end is a waste of your time. Maybe Matt can have a look at the logfiles to check if something catches his eye.
Thanks again!