Very interesting, but not surprising. It is often the case that facts can be distorted through oversimplification.
My recollections of the Teachings of the Master (Norm Abram) are that he never said that end-grain cannot be glued, but rather that he said that extra care had to be taken to get a strong end-grain joint compared with a long-grain joint. Specifically, he made a point of insuring that glue was applied to both surfaces, and where possible, he preferred using polyurethane glue because it penetrates into the fibers of the wood better than PVA glues (and note that Mr. Sullivan speculated that penetration of the glue into the end-grain fibers was a critical factor in determining joint strength).
However, I have a lingering question - does the length of the joint. matter? In Sullivan's tests, the joints all had the same length, but I wonder if the joint strength might increase if the joint length were longer - for the reason Sullivan explained, namely that the strength of the joint is limited by the strength of the lignin between the fibers of the wood parallel with the joint; I would like to see tests to determine if making that joint longer would increase the strength of the joint.