It doesn't matter what chemistry it is. If it's made right, and has zero defects in it, then it probably won't catch fire even if it's discharged all the way and hten recharged (albeit more slowly than normal, just to be sure). But it could, and there are some chemistries more likely to have such problems than others becuase of the way the chemistry works.
But if it's not made right, as is too frequently the likely case with cheap batteries like the RC LiPo and whatnot, or cheap ebay packs that may well be made from rejected cells from other places (no way to know for sure), then the risk of fire is greater.
What it comes down to is that it is more likely to be a risk under even normal conditions with a cheap pack made of unknown-quality and probably untested cells, than it is to be a risk under adverse conditions for a properly-made pack out of well-made and tested cells.
It is unfortunatley not possible in most cases ot know *exactly* what caused a pack fire, either because of the lack of remaining evidence to sift thru, or because of unknown conditions leading to it, etc.
Keep in mind that if you have some sort of cell-level monitoring system that prevents overdischarge of cells, you should be less likely to have such an event in the first place, and less likely to be at risk of a fire because of it. But it doesn't mean that some other defect or design flaw in the BMS or pack or wiring would not cause a problem.
That said, I have personally had some EIG brand NMC cells in my lighting pack brought down to 0V, and recovered them successfully (with trepidation at the time) to normal use. I sitll use those cells as part of my lighting pack, and they are at this moment powering my laptop via it's wall-adapter port.
I have never tested them (before or after) at high discharge rates, but they do give their full capacity as they did before, at the low rates I use them at. (1/4C or less) And they charge normally at similar rates.
I have never tried to recover RC LiPo cells in the same way simply because the QC on those is unknown and probably non-existent, unlike the EIG cells (which are used in larger EVs). I *have* deliberately overcharged such cells to see what they would do, and they can indeed vent and burst into flame under the more-than-extreme conditions I put them thru.
I have not tested EIG NMC cells the same way.
I have also accidentally tested NiMH, NiCd, and SLA under charge and discharge conditions that caused dramatic failures of the cells.
I'd intended to test a number of other types but I no longer have them to be able to do that.
There are some videos here on ES in one of member Liveforphysics' threads of various types of cells being THOROUGHLY abused with physical damage, overcharge, shorts, heat, and open flame, showing how they may fail under such unlikely-to-occur-under-normal-use situations.
http://www.endless-sphere.com/forums/viewtopic.php?f=14&t=17354
Under the right (wrong) conditions, any chemistry can fail dramatically.
