OP, thanks for this work but I have one suggestion that is important for accuracy here.
It is important to understand that filter manufacturer's only claim that their filter is xx% efficient at yy microns for certain filters in that family, not all filters in that family.
As you've seen with Purolator Boss, this is a misleading indication of overall performance and is often done for it's marketing advantages.
The Wix 99% efficient at 23 microns claim is a generic marketing statement that can be found in many places, but the 57055 that you have put against that specification does not list a specific efficiency (whereas Wix often does divulge it's beta ratios for individual filters).
Here is Wix's page for 57055: https://wixfilters.com/Lookup/PartDetails.aspx?Part=1870865
They only say that the nominal filtration is at 15 microns with no mention of efficiency and in any case this is not a measurement that is comparable to the result format that comes out of ISO-4548 testing.
My suggestion is as follows:
1) You should put in the filter model column the actual filter model number(s) that the manufacturer states were tested to arrive at the claimed filter efficiency numbers.
2) In the case of Wix (and maybe others), you can actually get hold of the filter efficiency for their versions of the filters that the competitors are basing their marketing claims of efficiency at yy microns.
3) So for Wix, I would suggest you do a cross reference to these competitor filters and put those numbers in as well. For the Fram comparison, you can take the average just as Fram do.
Example: Wix equivalent of Purolator Boss marketing claim
1) Purolator say the Boss is 99% efficient at 25 microns. This is based on PBL30001:
2) Wix lookup does not have a cross to PBL30001, but Fram do! It is TG8A:
3) Wix cross of TG8A is 51515:
4) And 51515 has a beta ratio 2/20=6/20. This is 50% at 6 microns and 95% at 20 microns.
The more like for like you can make your comparison, the more accurate it would be.
If you did this same exercise against Fram's designated filters, it would be very interesting to see the results.
nb: somewhat interestingly, Wix 51515 has nominal micron rating of 21, so the micron rating of 15 for the Wix 57055 for Subaru might make it a very efficient filter