There's also the possibility that the original team is deceiving themselves. By all appearances, they're deeply committed to this result (that could be an act, but that's not the sense I'm getting). But those biases could creep into the analysis of the material: in this particular field, a lot of things are witchcraft (witchcraft that occasionally works). So take a novel material with some interesting properties and human interpretative fallibility, and you get the paper. Which would still mean a powerful diamagnetic material, but that's... well, not something we'd have kilothreads on HN about.
My stance (did research in high-Tc superconductors in undergrad ages ago, but doing banal software dev nowadays): who knows. The "easy replication" idea has faded away, which isn't promising, but maybe? It's fun to put my faith in Twitter anime catgirl avatars, though, so I'm still rooting for it.
