If I was to guess, if there was an error, it would probably be the way the tester mounted the specific knife under the indenter on the anvil.
I saw a picture of a red manix2 IIRC under test (found it), and it was an assembled knife indented in the area between the G10 and the spyderhole.
It's hard to see much of the anvil, but it looks like there's just a pin coming up under the blade. Doesn't look like there's any other support for the blade.
Accuracy might be hit or miss, I'd imagine they try and get the result they expect, but if say a Lionsteel knife didn't lend itself to this method of indenting possibly due to handle shape or blade shape, you might get a movement of 0.001" or whatever.
I'd think some kind of measurement error is more likely, considering how they tested the Manix2, than Lionsteel heat treating to an intended 50 HRC. Let's be real.
Anyway, that's just a guess. I'd love it if Kurt would post some youtube videos of his test procedures. I'm by no means an expert, but I think others here might be.
Oh, here's an interesting comment by JLS who did a lot of Mule team hardness testing:
"...I still haven't found the test results for MT06 and MT10. If they're not in some old spreadsheets, they're lost and I don't know that I want to know the hardness bad enough to tear the handles off (once they're epoxied on, they stay that way...at least how I do it)."
You can draw your own conclusions how that might be relevant or not.
As I said, it's all just a guess.