There is a third option for capturing detail with multiresolution, and that is to simply place the Shrinkwrap modifer after multires and apply Shrinkwrap. The detail will then be captured in multires. I'm not sure how long this feature has been in Blender, but at least it works in version 4.3
But I did not know about the first method you demonstrated in the video (with a subdivided copy of the head), so I learned something new as well.
