Unless you are doing a lot of cave fills a tank should always be within the plastic range of elongation of the steel or aluminum.
We are seeing the problem of sustained load cracking on the old aluminum alloy. However, I have to question if it is really sustained load cracking or inter-grandular corrosion along the grain boundaries near the neck. There are a few videos of tanks being made that show how the neck area is formed from the pressed cylinder. This forming process is very similar swaging of aircraft aluminum control rods where we bring a larger diameter cylinder to a smaller diameter. In the process, the metal gets folder over onto itself and stretched. These fold lines can be a source of the cracking.
One of the many things I have against the eddy current testing systems that is now used is that it only checks the threads and not the neck area just below them where these folds may initiate the cracks. To do it properly you would need to use a medium sensitivity Fluorescent Penetrant and a black light. But then you would have to clean all of the tanks tested to remove the testing oils and the shops would refuse to do it.
Back to the tanks, a tank at rated pressure is well below the yield range and should be able to withstand that load indefinitely. Otherwise, all of those steel and aluminum pressurized storage tanks we see all over the place would be failing. Ever think about how many welding tanks are in garages and medical O2 storage tanks are out in elderly peoples homes and knocking around in their car trunks that are at pressure for years? Are they blowing up?
The one difference between a SCUBA tank and these other tanks is that a SCUBA tank gets wet and is subject to a higher corrosion potential. Corrosion pits are the enemy
The solution is to clean your tanks, dry them, and dont get them filled at a questionable dive shop or on a boat where water can enter into the tanks. The yearly visual inspection is the dive industries response to this higher potential. Also, dont store tanks on a concrete floor for a long time where they can get damp and stay damp for months. Put them on a piece of plywood or on a raised pallet so that they are on a dry surface.
Keep in mind that a single occurrence is not a trend, it is only a data point and you have to have all the information to understand that occurrence.
As for corrosion removing O2 from a cylinder, if you have any question, the check is very easy, just use a O2 meter as you would for checking NITROX.
If the reading is less then what is supposed to be in the tank, dump it and do a visual.
If it is the same as what is in the tank, dive it.
If you get a higher reading, well then try a new meter or recalibrate the one you are using.