They quote in the article that when moving in certain directions, the fermion's energy is completely derived from motion. So it's essentially taking the m out of E = mc², which is still neat, but not really something you can scale up