P60s have poor heat runaway so 2.1A should be kind of right.
3A sounds about right for an XPG2. Except that it’s in a P60… Well as long as you have lower modes, having the option of a turbo at 3A wouldn’t hurt 
Thanks for your opinions guys. Keep em coming :)
Yup, you guys are all right.
A copper MCPCB is great, but its full potential is countered by the P60 pill and host.
Some were building crazy throwers driven at 5.0-5.5A, but I think 2.0-2.5A would be optimal for you. (2.5-3.0A if you want to push it even further, but at diminishing returns.) Good luck!
+1. Even less possibly as the P60 has no thermal capabilities at all.
B.S.!
You can wrap your hand around the area of the P60 where the pill lies, that’s a great way to cool the light down.
There is a reason you have medium and low on flashlights… 
The LED wont sweat much before you find the light too warm to hold anyway as long as its copper mounted.
I think many people exaggerate the heat “issues” with the P60 format.
Until someone shows me some numbers comparing a P60 that has a copper mounted LED (and a copper ring between the pill and the body), compared to a similar sized light running 3 amps showing that the P60 are highly inferior in terms of lumen sag, im not buying it.
Running 3 amps in a P60, or P60 sized light is only a matter of how you use the light, and the environment you use it. With copper mounted LED, and maybe even a little copper ring between the P60 and the body, the main issue is in my opinion the surface area.
Put the light in a backpack, on high for 30 min in a closed, tight, environment and the body will reach 80+ degree. I would worry about the battery before I would worry about the LED, but that is me.
Use a 3amp P60 in the winter (around freezing point), some wind and it barely works as a good hand warmer when used for longer periods.
Running 2,8A in a Convoy S-series light on aluminium mcpcb, that I would question more, there is just not enough surface area to get rid of the heat and the mcpcb is a bottleneck that will let the LED run at higher temperature. The main problem with running 3A in a P60 for some time is the light is just not that large (not enough surface area to get rid of the heat), but who said you have to be able to have it tail standing for 10+ minutes indoor without the light very hot? There is a reason why you have medium, you dont need to use high ALL the time…
Im running 2,6A-3A on high in lights that are slighty smaller than P60. Usually 3A in P60 (sized), and often 3+ if its larger. Yet, I use high probably 90% of the time I use my lights… My point is, its mostly a matter of how you use your lights, and the environment you use it in. To some, 3A is perfect, to others its too much. I would not treat a copper mounted XP-G2 any less hard than a copper mounted XM-L2 as long as we are talking about a compact single cell light. Its not like you are going do damage the LED as long as its properly mounted.
A comparison between a Convoy M2 running 3A, and aluminium mcpcb vs a P60 with copper mounted LED, and a copper ring between the pill and body would be interesting to see.
People rarely say 2,8A is too much in Convoy M2, (or M1) with alu mcpcb, but 3A in a P60 with copper mounted led, then you are way over your head…
Even in small modified Maglites, shortie AA Maglite and other small lights, close to 3 amps seem to be fully accepted as long as there is a copper pill, but 3A in a larger P60 despite running copper mcpcb, then you are over your head according to the same people it often seems.
I use 3A high and copper mcpcb as default options for lights in that size. For me, and the way I use my lights, that is no issue.
To OP, if you are going to mod it, why not start with a 3A Qlite, and if you think its too much heat on high, and you want less light, just remove a 7135 or 2…
The XPG2 itself, on a copper board, will happily survive even direct drive from four Samsung INRs in parallel. I did that test a while back with 1-2-3-4 cells and posted the measured currents in a thread somewhere, but can't find it now...
No dobut…
Graps from Djozz… Sure, a much better environment than a P60, but still…
As I said in my long post, at 3A I would not worry about the LED as long as its properly mounted and attached.
I don’t worry about putting 3a through an alu xp-g2 now I’ve actually done it……just I’m on the wrong side of the diminishing returns curve, as said, got modes use modes.
Plastic P60 lights like SF G2 nitrolon, cheap zoomies, and smaller P60 host are not suitable for tuned emitters if you ask me.
I have 3A Vinhnguyen54 pill (sinkpad,solder filled) in probably the best P60 host flashlight in a world(Xsearcher) and it does not act as I would wanted to…
It will give 380 KCD for first 15 seconds, in next half minute or so it will drop to 350KCD, and eventually after 1-2 minutes on 335 and it will stay on this values.
So I could easily have 335KCD with 2,1A driver right?
I think with highly tuned emitters copper pill is must have… Only with it thermal path should drastically change…
Copper in aluminium body(cpu heatsink right?) should be excellent and more constant performer than brass pill with sinkpads/noctigons because after some time and because of poor thermal conductivity of brass most of heat will be hitting led sooner than you think… Thermal path would look something like 70% of heat on sinkpad/noctigon than 20% in brass pill, and final about 10% to aluminium body…
Copper star - brass pill - flashlight body
Pure copper pill should perform at least 100% better than brass one conducting heat instantly to the aluminum body…
So it should be about 60% in the copper pill and 40% in flashlight aluminium body
So it is only copper to aluminium thermal path…
I want to say that there is no point of tuning brass pill except for short bursts of light…
If the output drops but that drop isn't noticeable to the human eye, but does show up on a light meter, I have to ask - why not spend more time shining your lights at stuff and less time shining them at the damn meter?
I like measuring tools and hard data, but if the measured differences are meaningless, meaning not noticeable in the real world, you're just wasting time (and money, probably).
I have a P60 with XPG-R5 on alu board and copper tape wrapped driven at 1.4A. hot but not that hot, still ok.
So, having a higher output like say 2.1A for short time usage will be nice too
Thanks for all your input guys! Didn't think I'd get that many opinions on what I thought was a simple question.
Still haven't made up my mind. I have 2.1A and 2.45A Nanjg drivers here, but I'd like to try out the qlite as well. I'll probably change my mind seven times before the LED arrives, might as well put it in a convoy C8 
If you worried about the thermal pathway of a P60 setup take a look at my Noctigon mcpcb mod thread.
My XP-G2 @ 3A is no problem, you can have your cake and eat it too 
This was not on a P60, but I’ve been using an XP-G2 on a regular board from IS that in an STL-V6.
Still waiting for some XP-G2 on Noctigon from IOS, but I’m planning to replace the XP-G2, just because…
I’ve never had any issue with brass pills, Most mod kits come with them and they heat up pretty damned fast.
I also know that some have tested a copper pill in a c8 and measured no difference what so ever over a brass equivalent, the most important junction is the junction between emitter and star, which is why sinkpads and Noctigon’s have suddenly become so popular, not only for very high output lights but in more common builds, let the emitter work that bit more efficiently by getting rid of the immediate heat that much quicker.
To write brass pills off as useless is to totally miss one very important point, it is far far far easier to get a good electrical connection to brass, this is just as important as getting rid of heat, a high electrical resistance joint will suck output just as badly as thermal sag which is why I prefer brass pills to aluminium ones.
Tom E made me a p60 dropin with dedomed xpg2 at 3.5A on copper. 47kcd throw and 550lumen. Never had thermal issue since I use it sparely. It is just awesome!
any comparison beamshot with “standard” XPG drop in or before & after picture?