MCN published their original oil study in the August, October and November 2000 issues in response to reader and motorcycle manufacturer concerns over the API's then-new SJ performance requirements, especially the new "Energy Conserving" rating and the reduction in phosphorus to protect automotive catalytic systems.
On p. 26 of Part II (October, 2000) there is a chart titled "Viscosity Retention of Used Oil." They describe this chart as showing the results from conducting a Noack test on oil samples after they had seen at least 1,100 miles of use over two days in a Honda GL1500. The results are stated as the percentage retention of mass. The higher the percentage of retention, the more heat stable the oil is.
The Noack test MCN used heated the oil to 250° C for 60 minutes and had a constant flow of air over the sample. The idea is that the lighter elements of the oil will boil off, leaving only the heavier elements, resulting in both less oil volume and a thicker oil.
This was an irrelevant and misleading test for three reasons. First, the Noack test should have been on virgin oil, not used oil. We have no idea how hot the tested oils got during their use, and it is almost a certainty that there was some evaporative loss of the lighter parts of the oil during its use. Thus the different samples were subjected to different amounts of heat prior to the Noack test and we have no idea how much mass was lost initially. The results can not be used to accurately determine which oil is the most heat stable, as the best performer in this test might have already boiled off its lighter parts by the time MCN performed its Noack test.
Second, they did not measure viscosity. They measured the mass of the oil and assumed that mass equals viscosity. This is not true. Viscosity is a measure of fluid flow, not mass. So, the chart does not really show viscosity retention; it shows mass retention. Significantly, their test fails to measure the shearing action of the transmission and high engine RPMs.
Third, this chart tells us nothing about the viscosity of the oil after its use in the GL1500s. Stunningly, MCN does not provide viscosity measurements of the oil prior to its use in the GL1500s and after. So, we have no idea how well any of the oils retained their viscosity while in service. Oils that shear down to light 30-weight (or lighter) viscosities allow transmission gears to significantly wear and pit, as reported in SAE paper 961217. Thus is it important for a motorcycle oil to maintain a viscosity in the 40-weight range (or higher) to protect the transmission.
In conclusion, MCN's articles tell us absolutely nothing about how well an oil might hold its viscosity while in use.
There are other problems with the articles as well, like the assumption that more of a particular additive is better than less. This isn't true because the additives interact with each other and with the base oil in complex ways. Too much of a particular additive, in relation to the other additives present, can result in an oil that does not perform as well as another oil with lower, but better balanced, additives.
I love MCN and believe it is the best motorcycle publication available, but that doesn't mean its perfect. I compliment them for trying to tackle a difficult topic, but they did miss a few key aspects this time around.