It all boils down to: is something wrong with the steel, or something wrong with the JT show?
I told a materials scientist-type friend of mine about this, and he immediately said, "It's [the hardness] is banded? The top or bottom of a melt got used. It happens." That's what I think, bad part of a melt got rolled, or the melt cooled off too slowly and there was separation of elements, or more likely, not hot enough long enough. In short:
All the mass spectrometry in the world is, ultimately, just going to confirm the above in more detail. Unless the diminishing prospect that it's all JT's fault after all gets proven somehow.
Another point: when people new to knifemaking ask about what steel to use they are often told to use 1084 for various reasons, all related to the relative ease of using it, which could also be expressed as, "it's harder to fuck up." It also means that 1084 would be easier to pass along with more impurities or outright bad/counterfeit steel (talking about EBAY here), unless it gets sent to someone like JT who knows the difference and has the tools to test it. How many people used the same batch of "1084" to make some knives and never even knew anything was wrong with it? More than one, I bet.