While the theory works fine for static objects, how can it work for people? Did they have to precisely repeat their position while posing? Wouldn't it ruin the copying aspect?
He's using mannequins in the main picture at the top of the article, so I assume one technique would be to use mannequins as placeholders until you needed to do the distinctly human parts of the painting (hands, faces), then you could substitute a person into place and paint that part.