Well for me, these are a bust. Up at about 5-6 amps, turns blue, and lumens is about 850-900 at that point. The 4000K's looked warmer than 4000K (unless I mixed up the 4000K and 3000K, don't think so). I was using a black LK 5000 mAh cell, nothing crazy, but in a FET+1 modified Lumintop SD26 - probably 20 AWG wires, spring bypasses, mounted on a Maxtoch 26 mm MCPCB.
Took some pics, nothing special, but Photoshop is dead-spinning on me now so can't upload. Oops, good now:
I was drafting a long post earlier today but gave up on it. Figured I’d wait ’til they show up. I figured they’d act like the V2s and fall off a cliff at about 5.5A. These really are going to be well suited to triples. How’s the color consistency and tint? Green?
Yes, I did not have great hope that these are lumen/throw monsters (but 1000+ lumen would have been nice 
), but Lumileds is good in tints.
I guess a multi LED flooder of 5+ LED's on one cell would work. The tint looked pretty consistent but at the higher range of amps, it would get cooler, then higher still, turns blue. Guess a triple or quad would work with resistance.
Ran out of time for beamshots - snow on the ground, again, this morning and off to work. I'll double-check but the 4000K seemed pretty warm. I liked the consistency of the beam in a good thrower reflector, which the SD26 has.
Thanks for the feedback, guess the hunt continues.
Leds will not become much brighter, the evolution is close to its max, but there is still improvement on tints to be made.
(The new Cree XP-P hopefully outperforms the White Flat, but there is no way that the difference will be great. But it comes in 5000K which makes it stand out against the White Flat)
Edit: thanks for the link gchart, I should have done that.
The Cree XP-P? That’s The first off heard of that. Google doesn’t return much. Have any info to share?
Edit: I see Mark (from Cutter) shared about it and a XP-G4. I’m not so sure that picture of a XP-P isn’t really just a Olson picture. 
Not sure if anyone measured the tints, but the Boost HX looks far more neutral than a W1 white flat, though it does have a tinge of green. I like the HX way better than the W1 in tint. Looked at them side by side again last night. Tested them on the LT45 and the LT45 read the W1 lower, and the Boost HX higher than my LX1330B. This follows the pattern - LT45 reads CW's lower than the LX1330B, but higher for NW/WW.
TomE, do you have a bench supply to do current regulated tests? Could you measure a triple in comparison to a 219 or SST ? I have data collection and power supply equipment, but not a lumen tube. I thinks it’s time I change that… 
Plus replacing it might be easier since XP-L and XP-P share the same pads. I’ve chosen my KR1 with XP-L just for the nice tint, but this 5000K option could be ok as well.
I’m really okay with 3535/XP footprints and the next step up being 5050 for our application. Going down in size, like to the Cree XB ~2525 or further to Luxeon C’s 2020 (both have a neutral thermal pad) might be nice for keychain, AAA, AA formats or to offer more space on triple boards for extras. Otherwise 3.5, 5, and 7mm cover all the primary bases in flashlight world, IMO.
For modding though, it sure would be nice to have a consolidated footprint. Always sucks when your just an oddball mcpcb away from completing a build and there’s only one place to get it half way around the world.
I got a bench PS, but it's limited - not sure if 5 or 10 amps. It's the std budget one, like others have. I don't have a set up to drive a full up head (LEDs/optics) from the PS and record lumens though. Maybe simpler than I think, but never did it before.
I confirmed the LED I used was a 4000K, and in the lower amp range, the hotspot and corona are really much warmer, very yellow. Sometimes you get confused, staring at a beam, but putting it up against other single LED 4000K lights - big difference.
Also realized the SD26 has a FET+8+1 TA driver running Narsil, so I was able to change to mode sets and use a mode that was 100% of the 8*7135 bank. I measured 2.7 amps in this mode, and 670 lumens (calibrated). The good thing is the hotspot and corona are pretty even in tint, but very noticeably warmer than the flood area. I've seen worse for sure - I got Luxeon V's that have a blue ring in the corona, got other LED's with a blue spot in the center of the hotspot.
On max of the FET, I measured ~7.2 amps.
Also, now I noticed a dark spot in the center of the LED, so something burned up there. Probably too many times, and/or too long on the high amps, even though I've been doing it in as short bursts as possible.
I've got a 30v/10a, and it seemed going much beyond that started to get expensive pretty fast.
Anyway, a bit disappointing about the LEDs. SST-20s are nice if you can find a good tint bin (good luck), but the low Vf behind linear drivers makes me sad. 351d DOGFARTS has seemed really nice to me but I haven't had it in a reflector yet - and the high Vf puts it in XM-L2 territory for me: absolutely safe behind a FET, but it kind of stinks if your battery is low. Doesn't quite draw out the reds like my '20s, either.
Man it's no wonder people are still putting 219Bs everywhere, as old as they are...
Installed 3 of the HL2X 4000K's 80 CRI in a EDC18 (triple TIR) and it's looking much better! The beam doesn't have a yellow hotspot and corona, rather it looks very much like a typical 4000K triple TIR. On a not quite fully charged (but close) 35E, it read 10.2 amps on max and no signs of blue in the beam. While I worked on the EDC18, I installed the latest Anduril 2, and modded the switch LED wired to be controlled by Anduril 2. Stock, the switch LED is hard wired to be ON always. I'll record some #'s, but the output also looks decent. This is all good news for these low cost LED's.
Now I feel confident the 3000K's I have will also look good and performs well in TIR optics. I still got more TIR optic lights I'd like to swap out LEDs with. Also confident it should do ok on a 30Q at a bit higher amps.
Scheduled ship date for the 90CRI is April 21-23. I finally got my order today with the 4080s and some other stuff. Will be testing them out soon and try to post some data. These really are going to make most sense in triples and quads.
TomE, I’d love to hear how these compare inside the EDC18 lumens wise at top of ramp and turbo vs other options.
I’m probably going to throw one in a C8. I’m always looking for (or assembling) nice high CRI lights with good tint and a moderately throwy beam to use for nighttime nature hikes. I typically run them around 5-20 lumens. It’s amazing how far you can see down a trail with 5 lumens and a throwy beam on a moonless night.
Even at a max of 700-800 lumens, I’d say that’s plenty bright enough for a single emitter light for this purpose.
Okay… I should have known better to be excited and optimistic about these. I can honestly say I’ve never seen worse tint shift and piss green corona in a old era OP C8 reflector than with the HL2X. I’ve never really fooled with XPG3s because I heard about their issues before ever buying one (though there may be one or two in a junk bin somewhere from swaps). I will actually try and look for a G3 to compare the tint shift to. I’m about to try a slice/dedome and a triple in a TIR. I bought 10 and will sacrifice however many I need to characterize it and give them a chance at life, but I’m not holding my breath. I killed my first one trying to hot dedome it, but I did then get a measurement of the die underneath and heat tortured it until the wires de-soldered themselves.
Base chip was measured with my HF digital calipers to be 1.46 * 1.46 mm (+/- 0.02) . So 2.13mm²
I’ll attempt some pictures for y’all, too.
But thanks for all these tests.
Before I scraped the phosphor, I had about 2/3 of the dome off, and it looked way better. I’m headed back to fool with them some more now and also swap in the HHL532 in my A60