Sounds like he has taken some criticisms from here to heart. I love the way he places his temp probes in a ice bath (kinda like phaestus); but doesn't really describe what he did from that point on to actually calibrate the probes over any kind of statistical distribution. Its like if you put it in a ice bath, that alone does something.
Note also the TDX temps are a degree cooler...running 100Mhz faster with a higher V-core than the previous run. As a matter of fact, its the same temp as idle on the completely stock-clocked test. Calibration indeed.
I also find humor in his insistence on the word malleable being relevant here to the TDX. He even had a little hyperlink to a dictionary defintion of it. I guess from a technical point, he's right. Perhaps he doesn't understand the perception of the word in vernacular English though. Nice macro pics though; he must have a nice camera.