This entire topic is just beat to death because most folks are unwilling to agree/acknowledge the fact there is no 100% answer for all situations. When the question is flawed (poorly defined, unbounded ...), then resulting answers are just as bad.
The following things play into lube selection
- environment (temps, altitude, etc)
- OCI duration
- commitment (or lack thereof) to proper maintenance plan (not over-doing it, but not ignoring it either)
- operational factors
- contamination issues (fuel dilution, silica ingestion, etc)
- historical predominant issues specific to the engine series (fuel dilution, soot monster, etc)
Some engines are nearly impervious to grade selection; they just don't care. Other engines are somewhat sensitive to grade selection; the factors above can exacerbate the grade conundrum.
The reality is that while UOAs can help make this determination, the amount of time/data required to ACCURATELY make such a decision is woefully absent from any typical BITOG experiment. I have seen very, very few people here with enough commitment to be patient enough to properly establish averages and trends with truly accurate standard deviation. UOAs are great tools, but they are tools which are easily misused, misunderstood and misinterpreted.