There are reasons I stopped publishing comparative test results when I started selling products. The most important is that even if I could have been relied upon to try to be completely objective, it just isn't possible with a stake in the outcome. That's fundamental in any scientific endeavor. It's pretty clearly at play here.
Knowing what the temperature is and maintaining it is critical. With the oven readout and a single uncalibrated thermometer, we can only guess that the temperature was somewhere between 160°F and 220°F, with frequent cycling between the extremes. The samples need to be protected from the heating element when it is on. Steps should be taken to mitigate the temperature changes that inevitably occur in a standard oven. The fact that the sample plate cooled unevenly provides a good indication that it was heated unevenly. The samples should not have been touched until the entire plate had cooled to stable room temperature. Peeling them off when they are "cool enough to touch" is inaccurate and doesn't provide any usable information as a result.
Temperature is really important. 160°F-170° is what you can expect at the sheet metal when a vehicle is parked in the sun on a hot day. If the experiment had followed it's original design, it would have duplicated what the materials are exposed to in less than one summer day. Not sure why anyone would do that. 200°F-210°F is a better range. I've tested all of these products, except Audio Technix 80, for 8 hours at 200°F-210F and got very different results. If you heat a polymer above its intended operating temperature, you can get some idea of how it will perform over time.
The conclusion that P&S and FatMat are significantly different and that odor can be attributed to one and not the other in these conditions is puzzling.
Sliding after 8 hours is a huge problem, but the important information comes from examining the adhesive after the test. Here's where tester bias and not knowing what to look for are factors. We saw a wresting match with the Audio Technix products and saw how easy it was to remove the Second Skin products. There are two possibilities - adhesive failure, where the separation occurs between the adhesive and substrate (or constraining layer) and cohesive failure where separation occurs inside the adhesive. Adhesive failure, as long as it occurs above the required adhesive bond strength, is superior. Anyone familiar with products like these would say: "Wow, the Second Skin peeled off cleanly after 8 hours in the oven". Interestingly, high quality butyl adhesives gain bond strength over time. It's useful to test the performance of freshly applied samples against those that have been installed for a few days.
I'm not suggesting that the tester was being intentionally dishonest or that this test was staged to conceal a defect in his own products. What I am saying is that the test wasn't even close to being rigorous enough to reveal anything useful. The conclusions being drawn are incorrect - other than heating asphalt to its melting point, will melt it.
There's just no way around the fact that even with experience and the best of intentions, our brains are going to protect us from conclusions that work against our livelihood. It's as simple as that. When I see the tester's Web site meta title is "Deadener Better Than Dynamat" - a claim that is false in any qualitative way, it seems safe to view this "test" as an ad. Tests like these are sloppy enough when done by an unbiased party. When there isn't a methodology established beforehand and adhered to, sloppier still. That'd be fine for in house playing around to familiarize oneself with products. Does it belong in the public domain as proof of anything?