If you plan to develop this script further, I have a cool idea.
Rather than pulling the user’s input from a guide layer’s position value, is it possible to use
the output from an animation assistant effect(point control)?
In that case, there would need to be one controller for each vertex. This might be something the script could add to the layer for you with respect to num of vertices in a selected mask.
If you use the point control effects to pull position data, then users could apply tracking data to the point controllers, run the script and have that data applied to their mask.
Just my 2 cents.
Great work crgreen.