Sometime ago, at the outbreak of this controversy, I made my personal research on 123s dying unespectedly.
You can search among old threads. I found out that a 123 cell with a bad crimp seal, after being used for a while, will not hold residual charge due to the evaporation of an extremely volatile additive contained in the electrolite.
In my tests, this was brand-insensitive.
I want to emphasize that this is MY testing, not baked up from ANY source that I know... I believe, indeed, it is the real cause for failing 123s, and may be other types of lithium battery.
Just to refresh everybody minds with public data, a defective seal in a zinc-carbon cell produces evaporation of the water with an abnormal increase of the internal resistance of the cell; in the case of alkaline cells, it produces the conversions of the electrolyte, made of potassium hydroxide, in potassium carbonate (by the action of carbon dioxide contained in the air), first increasing the internal resistance of the cell and then deactivating it at all.
Now, lithium cells contains a component in the electrolite that is hundreds of time more volatile than water, so even a slightly defective crimp seal has very severe consequences on the cell functionality.
My 2 € cents
Anthony