1. The object is to find which medias pass oil more easily, and there is a correlation between visible media porosity/thickness/quantity and flow rate. Some medias pass oil very slowly in the gravity test. This is included as many people are concerned about flow rate in the smaller filters used these days, and want to have some idea of which will likely drop less pressure in use. In actuality, the filter is almost always relatively "invisible" to the pressure across the bearings, etc.
The relative results seem to bear this out weather I use thin (e.g. WD40) and pull it through under 10" Hg vacuum, or let gravity do the job.
2. Talcum powder is generally approximately 10 microns average.
Lately, I've been mixing some flour into the oil, too. Although this has not changed the ranking as to which medias filter more junk for me.
3. The proportional range is unknown. However the medias with the higher "factory" ratings certainly filter more than the less efficient medias, as you can see in the settling tubes.
This doesn't concern me, as with any home test, my goal was to find the medias (data unpublished or incomplete) which filtered out more of the same, given "dirt."
I experimented with different powders, because I needed to find some that would really show a difference...Too course and it all gets filtered, too fine and there might not be any difference to see.
The talc, and or flour seems to do well, as there is a considerable difference between what different filter medias let through, and a huge difference in what the sample tubes look like if a sample is drawn before passing it through the media.
Obviously, I can't give % efficiency numbers here, but a visual indication of different (often unknown) filter media that correlates well when known % media is compared to each other.
Some filter medias just let less junk though.
I looked at the Denso and K&N media, for example, both under the microscope, and in comparison of filtrate, (and ease of flow in the bench test) and find that...as good of filter as this might be...the media is quite open porosity compared to, say, a K&N of the same model. I chose the K&N myself, because it appears to filter better per pass, than the Denso for me. That's my own choice. There was no information published on the Denso.
I have gotten a lot of flack at times for this not being an ISO test with numbers, and I really wish I had kept my little basement playtime results to myself by now. But I thought some might find it useful in the vacuum of data.
Now I have a question for you:
What percentage increase in engine life, due to reduced abrasive wear, will you gain for each additional percentage over 95%, ISO mutipass, at 20 microns?
Obviously, it's an unanswerable. But rather than obsessing over numbers which do not indicate a knowable engine wear, I just wanted to know which filter would likey be the best filterers for my application. Especially, when the box just says "Excellent filtration!" and nothing more.