Your post below is quite likely the most devoid of technical knowledge that I've ever seen on this subject.
First off, the test itself is not applicable to an ICE. No reputable blender or manufacturer states that it is and ExxonMobil explicitly explains on their website that it is not an appropriate test for motor oils.
Second, even if you suspend reality and say that it does apply to an ICE the test PF does is not performed properly. The equipment is not a standard wear scar test machine. Therefore no repeatability and reproducibility data is available for this equipment. PF has zero knowledge as to what constitutes a statistically significant number of trials on any one test.
Third - and most damaging - the results are not analyzed properly, in fact they are not analyzed at all. The proper ASTM test for gear oils prescribes proper statistical analysis for results from a wear scar test, and PF (as well as that other well known website) perform no analysis of the data whatsoever. This in and of itself completely invalidates any consideration of this test. When the results of that other website's tests are properly analyzed it shows that there is
no statistical difference between any of the test results. This means the test cannot discriminate between any of the oils. Each oil tests statistically the same as every other oil. No researcher in their right mind would promulgate a test that is indiscriminate in the results like this one is.
As a result, yes Virginia the test is completely worthless and there is zero value. His tests are not standardized (they are exactly the opposite), they are not consistent, they are not repeatable, and they are dismal science. Precisely the opposite of everything you have claimed in your post.
PF's testing is valid - insofar as making a comparison of oils in the tests performed. His methods are standardized, consistent, fair, and repeatable; therefore decent science.
Where PF's tests may be good science, there is no evidence or proof of any kind that shows the tests' results transfer to engines. Sure, the cold pour test may have some relevance to cold start viscosity. Sure, the evaporative test may have some relevance to Noack. Sure, the lubricity test may have some relevance to engine wear. But without millions of dollars' worth of testing we will never know if there is a direct correlation between PF testing and real world applications.
What we do know is there is entertainment value there, and that the PF testing could have some bearing (engine reference intended) on real world performance.
Further, simply dismissing the testing as completely worthless is as much an error as saying the testing definitely correlates to real world.
And as far as putting money in his pocket? I say well done by Mr PF. He's found a niche market for his projects and gotten people interested in watching what he puts out. Applause for a successful entrepreneur.
(I also happen to think he is a member here, or has at least done a bunch of reading. We probably won't find out for sure.)