Maybe because they found all the oils in their 1996 test to be equal in quality, unlike the case in their much more limited test done about a decade earlier with SF oils, where they simply had new and used viscosity tested. In that test, they found that some 5W-30 oils and all but one 10W-40 oil dropped too much in viscosity. At that time, GM had complained to the oil industry about deficiencies in engine oils and said that none of the 10W-40 brands they tested met their quality standards.
Why bother with oil analysis when you do engine teardown inspections? I thought that oil analysis was used as a cheaper, faster alternative to teardowns.
You mentioned that 80% of engine wear occurs at cold start-up, but a maker of electric prelubrication pumps found that an engine equipped with a prelube pump had only 30% less overall wear than an engine that didn't use one. That's a lot lower figure than what makers of miracle oil additives have given. OTOH nobody has accused that pump maker of doing a bad test or scamming people.