Also, I should note that only modes 9 and 10 will work for duplicating smart objects. For modes 1-6 that need to re-align the images, those will create a new smart object when running. They will “steal” the layer mask and layer styles from the old smart object and apply those to the new smart object. So everything carries over in the layer itself….. except modes 1-6 won’t work for linking smart objects. For the image alignment and Face Detection, there was no other way to code it for modes 1-6 other than to create a new smart object.
However, modes 9 and 10 do work for duplicating smart objects. Those are the modes that you would typically do that for anyway, mainly to make print sheets.