The problem is that your sequence has already passed the “time check” prior to entering recovery and it won’t be checked again until the start of the next frame. I have modified this behavior such that after recovery is complete it will do a time check again before starting the exposure. This should fix the issue you encountered in the log you you posted.
The problem here is that the problem you are describing in words and the problem shown in the logs don’t match. You can see that at exactly 5:45, recovery stopped trying to recover (as expected):
[9/8/2015 5:45:00 AM] [DEBUG] [Recovery Sequence Thread] Recovery canceled… sequence has no light work left…
Your description of the problem indicates that recovery mode keeps trying to recover the sequence past the target end time (the logs attached to the first message show this is not true, but maybe there are 2 different things going on here).