Stimulating ideas.
My thinking is that the connection is made by the generation of a new layer of xylem, continuous across the joint and a new, continuous layer of phloem made by the mated cambium. I would think these would be no thicker than the layer thicknesses generated by the slower growing one, but maybe it is a one to many (as opposed to a one to one) connection. Still it seems that a graft with an older, slower thickening scion ought to succeed.
Ahhh ... Maybe it has to do with how pliant the tissues of the scion are. It would take some high precision equipment to make 'perfectly' flat cuts (I've never seen such equipment used to make grafts). Cut scion, insert scion, press and hold the flap to close the cut, then wrap tightly - soft scion tissues get flattened and in good contact with the stock's. Older stuff is too hard. Maybe?
Close enough for now. Get yourself a Shiner, if you like. I'm grabbing a Guinness.