I changed out my K&N filter for a Yamaha OEM. As promised, I'm testing the old filter to destruction, and verifying whether I can in fact actually detect any problems with design, materials, etc.
Visual inspection:
(can't get the image insert to work)
https://drive.google.com/open?id=1yaV0Xi0K90jfaelGTC6kTTMDsIEXZKi3
Filter still looks like new. No sign of the spot welds coming loose, or any bulging or tearing. To verify, I pried at the edges with a screwdriver, and also struck each side of the nut with a hammer. No sign of separation. Squeezing the filter walls in my hands as hard as I can, the metal does not flex or cave at all--the outer shell metal appears to be quite thick and strong. The seal is intact, with no nicks, tears, or abrasions.
Testing to destruction:
I then anchored the filter in a vise, and used my torque wrench with a 6-sided impact socket to verify how well the nut stood up to torque. I do NOT use the nut for installation, but I do use it to get the filter off (makes it very easy), and for safety wiring. I first tested at 10 ft-lbs of torque. No problem. I then tested at 20, 30 ft-lb. Still no problem, but it started slipping in the vise so I had to crank down enough to start to deform the filter. 40, 50, and 60 ft-lb, still no problem, but I had to keep tightening the filter in the vise. At 70 ft-lbs, the nut stripped so that I could no longer get any purchase on it with the socket.
https://drive.google.com/open?id=1KMSGeJMfJyeuUtx86Hw6RWyjDTAwrN0a
Even with the nut stripped and the filter partially caved in by the vise, the nut and plate on the end was solid, and not a single spot weld had failed. I then took it farther by using a screwdriver as a chisel with a hammer and forcing it through the spot welds. Some of them I couldn't get enough of a seam to enter, but I got through half of the welds on one side. Peeking underneath, it looks like there are additional weld points under the center nut.
Conclusion:
There was nothing wrong with design, materials, or workmanship of my filter. A sample of one, of course, does not prove anything about quality control across all filters, but it does seem to show that any problems with the end nut welds failing were specific to a particular batch, run, or manufacturing process and is not endemic to the entire line.