Well, frankly, I disagree with you CB.
UOAs are a great tool. They are not a perfect tool. But they are BY FAR the least expensive option to tract wear in a reasonable time frame.
Oil conditions are inputs to the equation. They are predictors. They can only imply that things are OK.
You can have a bearing going out, and yet the Vis, FP, base/acid, oxidation, fuel %, can all be OK. Nothing in the oil characteristics can predict abnormal wear of a component. These characteristics are precursors to supposition of how the fluid is aging.
Wear data are outputs of the equation. They are results. They tell us how things actually are doing in terms of component wear.
You can have a bearing going out, and the Cu and Pb will likely show an escalation of material shedding.
Admittedly, again, UOAs are not perfect. They do not see particles generally larger than 5um. But there's a reasonable ratio which exists in "normal" wear patters between what can be seen below 5um and what cannot be seen above 5um. There are plenty of SAE studies which indicate that the UOA data from spectral analysis correlates with other methods of wear tracking such as electron bombardment; component weight analysis, laser measurements, etc. Even PC studies on filtration have shown good correlation between particulate matter presence and UOA wear data.
Wear rates tell us a LOT. They are indicators of when wear is dropping, is steady, and is escalating. Quick, catastrophic events such as snapped con-rod are not going to be picked up in a UOA until it's after-the-fact. But wear related issues like bearings, piston bore scuffing, timing chain guides, etc can all be seen in UOA data. The issue is understanding the changes in wear rates, not just a static number. You have to know the means and standard deviations to determine "normalcy". Tools like statistical analysis allow us to really understand and focus on what is truly happening; what matters and what does not.
IOLMs are still notoriously conservative in their limits. They are certainly better than the old "3k miles; 5k miles" approach, but IOLMs only track inputs (temps, loads, duration, rpms, etc). They do nothing but predict a condition. They don't measure a result.
The inputs (oil characteristics such as FP, vis, ox, base/acid, etc) are all things to watch for; I do agree. But the reason to watch them is because when one of those conditions goes out of control, it's an indication that wear rates may change soon after. But those are not, in and of themselves, proof that wear is changing. Only wear metal data can prove wear is changing. We've seen countless examples of vis going "too thin" from VII breakdown, or "too thick" from oxidation saturation, and yet the wear rates for the OCI duration were perfectly normal. We've seen base/acid inversion, and nothing happened to wear. Etc ... The point I make is that inputs are a item to watch only in that they are a flag to indicate something MIGHT change in the output. When an input changes, it is not an automatic guarantee that an output will change; there is no assurance of correlation. When vis goes out of spec, it's not a reason to change oil for the sake of the vis; it's a reason to watch the wear rates with closer scrutiny because something MIGHT change. When an input shifts, it may or may not trigger an output change. Hence, vis, FP, fuel %, ox, etc are not reasons to change oil. They are reasons to watch for wear rate changes which might escalate to an undesirable level.
I like to make analogies; helps people get their head out of the topic and look at things from a different perspective.
What's more important to know about your favorite bastketball team?
The starting roster at tip-off (the inputs)? (who's on the court, what PPG and RPG do they average)
The score at the end of the game (the output)? (which team put more points on the board when the buzzer went off)
I think you know the answer ...
Outputs (results) are more telling than inputs (predictors).
The reality is that IOLMs and oil characteristics do not do a great job of predicting wear. Many times something can go "abnormal" in a condition, and yet wear goes totally unaffected. I have over 16,000 UOAs in my database covering a wide ranges of equipment and lubes and environments; I would know if this were untrue. Most of the time there is no correlation between oil conditions and wear rates.
Never confuse correlation for causation.
Without correlation there can be no causation.