This topic is predicated on a general misunderstanding of the concept of filtration.
Filtration does absolutely nothing to protect an engine; it is a tool to clean oil. It is clean oil that helps extend the lifecycle of the equipment. This is a concept of direct versus indirect relationships. It is the difference between causation and correlation. There are two ways to achieve a clean and healthy sump load adequate to sustain desired wear performance. One is to filter the fluid, the other is to change the fluid. Two means to the same end.
With that in mind, we can see great increases in both sump longevity and also in fiscal savings by using an appropriate filter. A decent filter properly spec'd for any particular application will keep the oil clean for a desired amount of time, and that is a huge savings. Look at two hypothetic possibility that are probably fairly close to reality. Persume these are healthy engines and of modern design .....
- Car #1 runs no filter, and has 2.5k mile OCIs, and can sustain acceptable wear. The 4 qrt sump migth cost $12 per OCI; that's $48 for 10k miles.
- Car #2 runs a decent $5 filter, and can run a 5k mile OCI, and still attain decent wear. That's only two sump exchanges and two filters over 10k miles; a cost of $34.
Hence, it's cheaper to OCI less often as the filter extends the OCI duration for any acceptable level of wear.
In fact, I would contend that today's modern equipment and fluids could easily do the entire 10k and still see wear-rates drop; a cost of $17. But that is becasuse both the fluid and the filter are working in concert, as designed.
A filter will not greatly alter wear rates in the front part of an OCI. That is because of two things:
1) a filter is less efficient towards the front of it's lifecycle
2) the contamination rate (generally fairly constant) really is not high enough to overwhelm the add-pack until much later in the lube lifecycle
Therefore, if you OCI frequently enough, you see practically no difference in wear in shorter OCIs, with or without a filter. Only when you extend an OCI would filtration make significant shifts in wear rates that the oil would otherwise be overwhelmed.
If you buy/read SAE 2007-01-4133, you can see that using a decent filter (nothing super premium, just a normal filter) over an OCI stretching out to 15k miles has the wear rates going down. But the majority of wear is right after the OCI. Then the rates trend down as the tribochemical barriers are re-established. It is that chemical-physical film barrier that has the greatest affect on wear. Until it is upset, it just continues to get "better" for a long time.
The contamination of a fresh sump load is not nearly high enough cause horrid wear at the front of an OCI; it has to accumulate over time. So the filter is there to prolong that steady-state. But it really only becomes effective past the point that the oil would otherwise be overwhelmed. What actually bumps the wear rate up at the front of an OCI is the detergent/dispersent chemistry attacking the existing tribo-chemical barrier. After that is re-established and settles, the wear drops back down and continues down even as long as 15k miles. That is borne out not only in the SAE article, but also in my UOA macro-analysis database information as seen in my "normalcy" article.
In both micro and macro analysis, we can see that wear rates are not generally affected by using either standard or super-efficient filters in "normal" applications. So to some degree, the level of filter efficiency is not having an effect on the wear rates. That means that somewhere, before the efficiency does matter, there is some level of contamination that is incidental and inconsequential to the rate of wear. I have yet to see evidence that super-premium filtration makes a significant difference in "normal" applications (I exclude obvious situations where only "non-normal" can apply; I define "normal as short-to moderate OCIs with a healthy engine and typical use).
Don't construe this to mean filters are not important; they most certainly are. But you have to understand the relationship of direct causation and indirect correlation, and not only how but WHEN these come into play.
The synopsis?
At some low OCI duration, filters do very little.
At moderate OCI duration, filtration is important, but efficiency has seemingly no effect as long as a minimum threshold is achieved.
At long OCI durations, filtration makes a significant differnce (perhaps capacity even over ultimate efficency).
So to the OP, you question is "what is actually important"?
Well - that depends upon your OCI duration and the general healthy of your equipment.
