Unfortunately, I think the green tagging box is pretty detached from having anything to do with what started the event, or even whether the motion identified was within your detection window. We had a lot of discussions about this in the early days.
Motion Tagging sensitivity is fixed (let’s say 50%?), and not related to the user-determined Motion Detection sensitivity. It is always set to that level, and applies to the whole screen regardless of detection zone. Also, when there are multiple moving objects, the largest moving object will be tagged.
So on a given video, you may or may not see the green motion tag, and any tag shown may or may not be in a detection zone area. It may or may not have anything to do with what triggered the recording. It is simply an indication of something that exceeded the pre-set and unchangeable motion sensitivity. This causes a lot of confusion for users, who try to connect the tags with event captures. They aren’t directly related.
But I agree, there should be a connection. They should trigger at whatever sensitivity threshold you have set, and only if motion falls within your selected detection zone. This makes much more sense, and allows you to infer things from the tagging. 
This was the only Wishlist item I noted at a glace that seems to try to address this issue: