I neglected to mention one discussion point in my OP--
Is it possible that a leaky Ultra could outperform a nonleaky filter with lesser media? Certainly it's possible, but it depends on what you are comparing.
An Ultra that fails the "flashlight test" still has a ton more flow area through the media than it does through the endplate gap. Think of the relative cross sectional flow area (effective orifice size) of the "defect gap" vs the media. Obviously the media is MUCH larger a flow area. The media has to accrue a fair bit of restriction before an appreciable fraction of flow prefers to go through the leak path vs the media.
For sake of illustration and exploration, let's say that a leaky Ultra starts out with 5% of the flow going through the leaky endplate when new and rises as high as 20% as the filter approaches the end of its life.
How much actual efficiency is lost? Well, the Ultra is supposedly a 99% at 20 micron, so that's a ß100 performance at >20 micron. But if only 95% of the flow is passing through it, the effective Beta is only ß95@20µ at beginning of life. And at end of life it would drop to ß75@20µ.
Compare that do a Micro guard Select that is ß100 (99% efficient) at 25µ.
Because Beta ratio changes so much with particle cutoff, it's almost certain that a leaky Ultra at 20 micron beats a non leaky Microguard at 25 micron. Even at the end of the Ultra's life where it's lost efficiency, it's almost certainly still better than a brand new non-leak MG Select.
This is because the Ultra being rated way down at 20 microns with a ß100 means that when you rate the Ultra at the 25 microns that the MG Select is rated, the Ultra is likely to be closer to ß1000 or even higher. Let's conservatively say it only rises to ß500 at 25 microns.
If the Ultra's media is effectively ß500 at 25µ vs the ß100 of the Microguard Select, that means the Ultra could internally leak 80% of the flow and still be just as good as the Microguard Select.
Now I haven't seen all the "flashlight test" photos floating around, but I've yet to see any gap large enough to pass anything close to 80% of the total oil pump outlet flow when the proportionally large surface area of the media is available.
Yes, it's bad that the Ultras leak internally. But it's not at all so bad that it's worth going to a much lesser media to try and get away from.
Internal leakage isn't great, but it has to be pretty bad to give up running some of the best media available in many applications.
