hehe.... we do this every time. How many times have we disputed the value of the "cheap" cells FallingWater?
Ok... I'll give it another go...
We have in fact had a li-ion explosion in a 4xRCR123 powered Kai light that was being used with cheap no-name unprotected cells. What's amazing is that it's VERY RARE for Li-Ion cells to develop problems during a discharge, they almost always fail while being charged. point being, if they managed to make a Li-Ion cell so unstable and so out of whack that it failed on a discharge, then the likelihood of a vent during charging is in my eyes, substantially higher.
If someone takes shipment on a cell that has been resting in an over-discharged state for potentially years, they are putting themselves at risk even trying to use the cells, many around here have gone ahead and charged them up, only to find substantially lower than label capacity, the question then becomes whether the battery is suffering from low capacity because that is simply how it is out of the package, or if it is actually that severely aged fresh out of the package... Considering the number of these cells that won't even hold a charge over 4.10V new from the distributor, It's very reasonable indeed for me to point out that these are risky cells that are much MORE likely to vent-with-flame than a more reputable brand.
saying that problematic cells are rare is not really accurate, here's just a few examples from around here. Keep in mind that AW has been selling large quantities of cells to CPFers for a much longer period of time, and overall I think we still have less incidents with AW cells than we do with the ultra-super-trust-fire cells which are a relatively new thing on the market.
Here's a 900mAH rated 14500 trustfire cell that only lasted 20 minutes on a 0.94A load, Meaning it came in at less than half it's label capacity. I realize inflated ratings are common in all brands, but this is more than double, which can be very risky when trying to calculate even a ballpark maximum safe discharge rate.
http://candlepowerforums.com/vb/showthread.php?t=186042&highlight=trustfire
Here's an example of a failed protection circuit and 18650s coming in at about half their label capacity. Major heating problems during charging/discharging also.
http://www.candlepowerforums.com/vb/showthread.php?t=173760
Here's an example of more misbehaving cells, one of them won't even get above 3.5V, the risk of an incident with that cell was pretty high IMO.
http://www.candlepowerforums.com/vb/showthread.php?t=179045
here's something interesting, we have 2 very different specifications for what appears to be the same cell... now what do we know?!
http://candlepowerforums.com/vb/showthread.php?t=173341&highlight=trustfire
protection circuit on cheap RCR123 failed:
http://candlepowerforums.com/vb/showthread.php?t=181958&highlight=RCR123
Here's an RCR123 explosion with cheap unprotected cells:
http://candlepowerforums.com/vb/showthread.php?t=180597&highlight=RCR123
Luckily as I understand it, nobody was hurt, no fires were started, it was outdoors and the flashlight was tossed in snow which prevented a potentially worse scenario, Kai replaced everything if I understand correctly. So there's good an bad there.
Here's a strange problem, cell did get some good use for about a year before failing, but this one was borderline explosion, heated up to a very dangerous point.
http://candlepowerforums.com/vb/showthread.php?t=186165&highlight=ultrafire
Here's an example of some brand new 18650s that refuse to take a full charge and perform poorly compare to aged Panisonic cells. These cells will barely hold above 4.0V, suggesting that they are already nearly completely "spent" as far as cycle life is concerned.
http://candlepowerforums.com/vb/showthread.php?t=184062&highlight=ultrafire
in this thread:
http://candlepowerforums.com/vb/showthread.php?t=182643&highlight=ultrafire
LEDite points out that he has had hundreds of cheaper cells fail prematurely. Then 2 responses down, forum Member Mash points out that he has had a "dud" 18650 from DX.
Here's 2 out of 3 cells shipped dangerously over-discharged.
http://candlepowerforums.com/vb/showthread.php?t=180505&highlight=ultrafire
here's a 2xRCR123 setup where one cheap cell discharged to 0.8V while the other was still 3.8V, obviously this was also probably user error in conjunction with the mistake of buying unprotected cells, but the fact that the cells were so inconsistent capacity wise can be a major problem.... this is the type of situation that could lead to a cell be reverse charged and going haybonkers.
http://candlepowerforums.com/vb/showthread.php?t=176800&highlight=ultrafire
http://www.candlepowerforums.com/vb/showthread.php?t=169200
in that thread, you can get a better idea what you are getting for your money with the AW cell, his latest revision of the protected cells includes great reinforcements to protect the PCB, internal PTC as a backup measure, and industrial grade shrink-wrapping that in my experience is in fact very durable in compared to other cells. When you buy an AW cell now you are getting Pila level quality (possibly better in some ways) for half the price. That is great value and piece of mind.