FWIW, and without any intention of claiming that this is the cause of the Aleph incident in question here, I share the following anecdotes:
Ever since reading Newbie's results, I have been testing every battery on a ZTS from Lighthound to be sure it had the charge I expected. I have been VERY careful with multi-cell lights, and only use batteries that show 100%. I run batteries in multi-cell lights down to only about halfway, at which point they go into single cell lights.
Maybe a year ago, I ordered one batch of 36 Battery Station batteries, and they ranged from 40% to 100% on my voltage meter when new! Doing the math there puts you into a possible danger zone less than halfway through the anticipated cycle. Needless to say, this is the last time I tried to save money on batteries. Cheap brakes and tires, anyone?
[Insert comment: I am pleased to see that Battery Station has responded to the problems with changes in the design of their CR123A. I think I will respond to this positive news by giving their new product a chance and ordering a test batch. Personally, I wish them all success with a positive rollout of an improved product, and am impressed with this responsiveness.]
I have checked over 100 new SureFire batteries and every single one has read 100%. No exceptions. I check each battery at least 3 times and sort them into boxes by percentage (I sort used ones from the multi-cell lights too). The meter varies widely in what it reports, so I keep checking until I get a few consistent readings in a row.
This raises something of a gray area for me. If I buy off-brand batteries, do not check them, get mixed voltages in a pair of new batteries, and have a reverse charging incident, is it user error? Ostensibly not, but it's arguably a case a dedicated user can avoid. Should this trouble be necessary? Definitely not. My experience has led me to have a high degree of faith in SureFire batteries, but do I trust their track record enough to stop checking? No.
The one thing these anecdotes of mine can conclusively add to this thread is that relying on batch numbers to conclude that they were "matching cells" is a fallacy.
I think these lights and batteries are safe the way we use them. But I don't intend to be the 1% (or probably much, much less) who has a problem if there is anything I can do to minimize my risk. Safety first!