So here's my opinion of the BYZ/TUR situation. I should say beforehand though, that I don't mean to step on anyone's toes, and mean no offense; just straight up opinion.
Alright, so I've looked at the 4 events...why is there an AI event? I don't see any reason to have one. I can see reason for an automatic seceding event, but with perhaps more triggers (for instance, why should TUR get Thrace if a country other than BYZ ownes it? - read on before you rush to answer that). I'm not prepared to vote to give a carte blanche in this situation to any tag, regardless of anything, that would no wbe able to hand Thrace over to TUR based soley on TUR controlership. Not just yet anyways.
One would think that including the times that BYZ actually keeps Thrace, this should boost the overall probability of TUR acquiring Thrace by the CoT deadline to acceptable levels. At the very least, we attempt the light handed approach first, and this fails, we then have better reason to go to the heavier solutions.
Also, there was a mention about scripting an event in case of the BYZ/TUR vassalage being broken early (a few posts after doktarr's post #88). That would be simple enough to do, and I think that's a good idea (I believe it may have been doktarr's but I'm not sure). At any rate, you'd post an interpretation for what that vassalage meant historically, then use this for the basis of formulating an plausible ahistorical implication/consequence of what it would then mean to have such a vassalage broken prematurely. Finally, you'd translate such consequences into the action commands of an event that is based on no existing vassalage between BYZ and TUR (maybe use a relation and stability trigger in there?).
I'd guess that this would probably give TUR grounds to DoW BYZ, though the event should not, IMO, force a DoW (that'd be unfair to the BYZ player) but rather, the A choice should lower relations terribly, provide a casus belli (if not a core, if there isn't already one there) and maybe even some troops and gold for the influx of hawks at Court, eager to teach BYZ a lesson in submission. B choice would be to let the issue drop.
Perhaps, if doktarr considers the above suggestions, we might be able to hammer out a compromise set of events. Again I stress they're only suggestions.