My K2 cells performed very well (capacity and discharge curve under load) against their factory specs when I tested them new(with an Opus 3400) [which I do not use to charge those cells, BTW], and in actual use I haven't noted any issues. That said, they've had a fairly easy life without too many charge cycles (they aren't in my more heavily-used lights), and I've not yet done followup testing on them, so I can't say how well they've held up performance-wise in terms of cycle life. I plan to re-test them, but haven't gotten around to it. If you noted decreased performance over their life, I guess that means I'd better do that re-testing to see how mine perform now. If they've degraded significantly I'll be disappointed, because I don't have another good alternative to using CR123A primaries in those lights. BTW, I assume you were charging the LFPs you were using to the proper 3.6V with a charger supporting that, and NOT to 4.2V like most other lithium ion chemistries use. If you charged LFPs to 4.2V, that would greatly decrease their performance / operating life. You're probably aware of that, but just wanted to mention it to be sure, as they don't like that.
As far as your question on whether or not they're "significantly better than standard lithium batteries", if you mean standard lithium primary CR123As, the only significant advantage I see is that they are rechargeable (with a claimed high cycle life), albeit at the expense of lower capacity, while still performing well otherwise (fairly flat discharge curve - actually flatter than a CR123A primary). Their rechargeability is really the only big 'plus' over 123 primaries, as it saves a lot in operating cost vs primaries, and the LFPs really are the only good alternative I'm aware of for a CR123A. In the case of the PT 1L-1AA however, you DO have another rechargeable option, that being the Eneloop AAs, and you may consider them a better option than the LFPs since your light can use AAs - although the output level is undoubtedly lower with Eneloop AAs than it would be with the LFPs, which if correct would be another 'plus' for the LFPs.
As for the O volt situation with the Olight 16340 cells: I suspect that the 0 volt condition is due to the protection circuit having 'tripped', and not being heavily discharged. If you put them on a charger that resets that circuit / condition and then begins charging them, I suspect that if you stop the charging immediately and check them with your meter, you'll probably find that their SOC is not actually '0 volts', but actually much higher. I don't think they're being fully discharged, but rather just 'tripped'.
I have had fairly crummy results with the Olight 16340s, had a lot of nuisance trips under load (like I think you're seeing), and discontinued using them. Conversely, I've had very good results with the Fenix 16340s, and have quite a few in use.
Don't forget that you MAY have damaged that light by attempting to use the unsupported 16340 cells. I don't know, and I would never even try them in that light to try to find out, because the risk of damaging the light is too high. If the light still works correctly / to spec with recommended cell types at this point, then I guess you got very lucky - but I wouldn't recommend continuing that practice.
