Right after djozz’s post in that other thread, is this claim from eebowler:
So, if that is true, it was done at least 10 years ago by someone not on BLF at the time. 
If It’s just to make the cross to disappear why you don’t just sanded the top of the diode?It will allow you to get little more throw?
I was worried about the domes on a couple of mine after cleaning the solder pads and wires with some isopropyl, they’ve gone slightly opaque which I didn’t expect with the isopropyl.
I’ve been posting none stop today, nothing else to do. :weary: 
How much current are you driving it at? I would personally add a note of caution to this.
………………….
Instance #1:
In my XHP70 DST, I also wanted to get rid of as much of the artifacts/donut as I could, but I didn’t want to mess with my beautiful smooth reflector (in case I want to go back to thrower later). First I shaved/sliced down the dome (the dome magnifies the emitter, and it also magnifies the gap between the dies). Then I very lightly sanded the flat top with 800 grit wet/dry, and cleaned it off with some isopropyl. It ended up with a very nice frosted effect. I fired it up with a FET driver at around 8amps, and it was almost ruined. Apparently the frosting caused a lot of internal reflection resulting in heat. Read more about it here: Just toasted an XHP70... How can I avoid doing it again?
Instance #2:
When I was building my triple XHP-50 Fandyfire G36, I wanted the same things as my DST, so I shaved the emitters down again. This time I did not sand them. On one of them my shave job wasn’t great. It was a little uneven, but still for the most part clear so I figured it would be fine. Also driven by a FET, it smoked out just like the sanded XHP70 did. This time it killed two of the 4 dies in the process.
I'll second what pilotdog68 posted above.
I've burned up an XHP50 that was shaved down, when driven DD. At first I thought there may have been a foreign object left on there, but when I picked it apart, it looked as though it had burned from the inside-out.
Sanded mine with 400grit, it really smoothed the beam. There was a slightly visible doughnut hole and now it’s gone. It is driven by a FET zener modded and draws 6A+ on turbo with 2x efest 18350. Ran for 30 seconds and seems to be fine. However, output dropped by 8% when tested ceiling bounce vs untouched dome, but still produces more light than a triple XP-L setup at 9.6A.
Ok, you guys had me really concerned for a minute or five. I ran the light (3C mag) at 7-8 amps for around three minutes on king kongs and for about an additional minute plus at 9.5 amps on powerizers. Had to quit because I couldn't hold the light anymore due to heat. There was no tint shifting and no visual signs of change in the dome. I think I'm alright with this one. I wonder if the damage sustained by Pilotdog's emitter was just having the frosted effect that much closer to the dies due to the dome shaving?
Maybe you’re right about leaving the dome on vs shaving, idk. In my DST it went in the first 2min. It looked great at first, then slowly the tint started to shift a bit…then the beam started to turn a weird shape. I shut it off and the reflector cone was already completely full of smoke.
That's a great looking beam, I'd trade an 8% loss for a perfect beam anytime.
I have 12 lights running Shaved and polished XHP70’s and 50’s and only had 1 start to burn after shimming the reflector for better throw and I know something fell onto it. It ran for 3-4 months with no problem. After re-sanding and cleaning it vigorously with Dawn ultra and cotton pads, then blowing off with compressed air, its working to this day. If it had burned from the inside out I don’t think it could have been saved.
SD75 Shaved/Polished XHP70 FETDD 12.6 amps. Smooth Reflector (not OP)
Are you shaving the dome in steps or all at once. I’ve sliced and sanded 4 XHP50s by making multiple slices and sanding ever so lightly as to not physically stress the LED and have no problems running good 18650s DD. My last one I sliced but didn’t sand and there’s no difference in beam profile with OP.
I built this on April 6th, 2015, direct drive on a Zener modified FET driver
I’m not sure how much collective time it has run, but it does pull 13A on a pair of 18350’s and 19.1A on a pair of 18650’s. The shaved XHP-50’s sit under a Ledil CUTE-4 optic. Good for 4896 lumens with the half cells, 6037 lumens with full cells. Only tested with full cells, it is used with the little Efest Purple 18350’s. This is a LuckySun D80.
I looked back on my build thread for that light, & memory is playing tricks on me. I'd have to say it was my fault for not cleaning the shaved surface properly. I did run the light for a while with non-high drain cells, but as soon as I dropped a couple of 25R's in there it burned up straight away.
What was left of the dome was charcoal, right down to the phosphor :(
From this; To this;
Oops! I spent the afternoon making an extension tube for my D80 Quad XHP-50. Got it all fixed up, charged up a couple of Efest 35A cells, and got right at 6000 lumens shortly before it dropped hard and started turning blue. Yikes! Too much cell. lol
Yikes indeed. I'm building a quad XHP-50 now and I'm hoping I don't run into the same heat issues as you. My host is bigger but I want to run it off high drain 26650s.
kyfishguy, keep in mind that it’s eXtreme High Performance, XHP.
I have 2 copper 25mm mcpcb’s underneath the 4 10mm SinkPADs (masking sanded off for the copper discs) but there’s just not enough mass to take this much heat fast enough to do any good. I’m thinking about rebuilding the D80’s head with a head filling copper core that screws on like the bezel, extends up and out and the bezel will then screw into the top of this copper. Put the quad back together inside it, should be good to go. Maybe 12 ounces of copper or so with some of it open air and the rest counting on the fins to pull heat away.
My FET driver will give 22 amps to these 4, they should be able to take that fine, just need the heat taken away.
Sooo, it seems it may be confession time. The sanded XHP-70 in the maglite is running strong with no hint of degradation or other harm to the dome. The four XHP 50's I sanded went into a quad light (build thread in the near future). I finished the light around midnight and rushed to test it out. Amp draw on a pair of powerizer 26650s was 15 amps.
After amazing myself with the wall of light for a minute or two I went back inside and happened to look at the domes. Two of the domes had black carbonized tunnels in them from the apex almost down to the phosphor. They like very symmetrical mini volcanoes. The third dome had cracks in the surface like old leather and the fourth was perfect. Yes, I completely forgot to wash down the domes after sanding. I expect some particles from the sandpaper remained on some of the domes and burned their way down into the dome material.
After care included dome shaving and polishing for all four emitters followed by thorough cleaning with alcohol. I probably shouldn't have used alcohol but aside from some smoking for a while on turbo I seem to have gotten away with it. The beam is still clear of artifacts with OP reflectors and much more pleasing than the domed emitters were with the reflectors I'm using. I doubt any more dome sanding is in my future since the shaving seems pretty easy and is perhaps a bit less risky.
Saved and shaved!!!
How much does the tint change from the dedoming? I am hesitant to try it on my K60, Would be great if Cree made an XHP-70 “HI” and put the dies closer together like the XHP-35…
The tint shift is pretty minor, just a hint of extra warmth in the general direction of yellow in the 2B tints I shaved. Pretty nice really.