You have presented BITOG (or crowd sourcing) as an effective means of macro analysis and that is simply not true.
I'm honestly not sure what you mean by this, so I'm asking not to argue but to understand.
As a clarification, I'm not suggesting or implying that BITOG data (in it's raw form from the site) is a means to compare/contrast it as macro data. If it came across this way then I apologize to all because that's not what I meant. What I'm saying is that if you have large amounts of credible data, analyzed with proper skills, that macro data can provide a basis to singularly compare/contrast one's UOAs against the statistical normals. One can use the macro data basis to understand if their own personal UOAs are "normal" or abnormal.
To understand what is normal to a single example (one engine for example), you must do micro analysis.
To understand if that single example is responding in a typical manner against the masses, macro data is very useful.
It is completely improper to take one's single engine UOA and compare it to another persons single engine UOA and make any decree of what's better or worse; that is patently bad methodology.
In this specific thread, I have over 600 UOAs for the 4.6L engines, from all walks of life, with all manner of lubes and filters. We can easily understand what is "normal" across the total application life presents. Knowing this data allows us to understand what is "normal" for that engine series in all environments, using a broad base of lubes and filters, for varying OCI durations. With that much data, I can even narrow it down to OCI durations and some of the more popular lube grades/filters.
What I am completely confident of is that for most typical OCIs (out to 10k miles), the lube grade simply does not affect wear rates; the amount of standard deviation due to typical "life" use is greater than the delineation of lube grade for this engine series (and many others, for that matter). This is driven by gobs of credible macro data. To know if lube grade has an influence on wear rates for a specific single engine, micro data analysis would need to take place, and that takes FAR more time/money than anyone here has in their calendar/wallet.
Does that clarify what I meant? If so, how does that relate to your objection?