I bypassed the springs, but am still using stock cells, which are probably not high-current. I think I gained something like 10-15% turbo output from the bypass.
Still, Iām glad to read your input; better late than never! 
That is what I expected (and also already found out by now).
I might be willing (and capable of) to switch out the FETās for higher quality, but I havenāt got the knowlegde to dertermine what would be suitable replacements and their quality. So thatās a step Iām a bit hesitant for.
Iām also thinking of doing a spring bypass for a while now and maybe even also swapping the wires to the LEDās for thicker ones, but so far Iāve been held back because Iām not fully confident if the FETs (and/or other driver components) will hold up. Last think I want is to have take the whole flashlight apart again and (way) more extra time in it. Ideally I would like to take the flashlight apart just one time! 
Iām currently running 3x Samsung 40T(5) cells in it, but with having Ampace JP40 or Eve 40PL/50PL available for really good prices now, I would gladly swap to those cell. But also having read stories from people whose Q8 Plus (apparently) already failed by putting in Molicel P42A or P45B cell, Iām not sure putting in potentially even more powerful cells (like JP40/40PL) is the best idea, without doing part swaps first as a precaution.
1 Thank
In my opinion the BSC009NE2LS5ATMA1 is probably the very best option.
DigiKey:
https://www.digikey.com/en/products/detail/infineon-technologies/BSC004NE2LS5ATMA1/13919216
1 Thank
Thanks for the good suggestion, I will remind it.
Downside of both Mouser and Digikey are the high shipping fees (when ordering below ā¬ 50,-) of ā¬ 18,- and ā¬ 20,- respectively. Both are not stores where I would normally hit an order amount of ā¬ 50,- easily, so it might take me some time to collect enough items to make it worth/viable to place an order!
1 Thank
So good to hear from you, thanks a lot for the info! Iād also like to do a FET swap at some point in the future, but am worried that my clumsiness would mess up some other part of the board. Iāve only done reflows of MCPCBs, not fiberglass PCBs. Maybe when I become more skilled at electronics workā¦
Itās a miracle that your springs are holding up to the 45T cells. Iāve heard lots of reports of high-current cells permanently deforming unbypassed springs from heat. Even the stock low current cells get the springs extremely hot when I opened up the light immediately after turbo to check. And once the spring is bypassed, the FET would be the next possible source of failure.
This post makes me a bit concerned about a LHP531 swap: they have higher Vf than XHP50.2 on paper, but recent test of LHP73B vs XHP70.3 strongly suggests that the LHP will have much lower Vf in reality. But for now Iām probably safe with the stock batteries.
I feel your pain trying to order small quantities from these stores. My university makerspace is restocking some electronics, so Iām hoping to see some suitable MOSFETs, if Iām lucky.
Iāve made a stupid mistake: it are Samsung 40T(5), not the 45T. The 40Tās are still decent high drain cells, but the (newer) 45T do outperform them though. With the 40Tās (still almost as new) I havenāt got issues with the springs so far, but if I were to put in really mighty cells as the JP40 or 40/50PL, Iām not so sure that would still be the case! 
2 Thanks