Even if the XP-L stays within 10% of it's output when de-doming, with a triple light that's a 30% loss in output! With a hot rodded triple, that can be a very significant lumens loss.
Say we're seeing 1400 lumens per emitter for 4200 OTF lumens with the domes on. Take the domes off and lose a mere 10% and we have a 520 lumen loss! Yeah, I'm all for keeping the domes where they belong, on the emitter instead of on top of my computer. (I have a little collection of hot de-domes. :) )
I used a cordless Makita hand drill with variable speed and did most of it at very low revolution. Then I set it to high speed mode and ran it fast to sort of polish the cut. Works so so. There was a bit of slag along the cut edge that I scraped off with the razor sharp tip of my Sage 2. Now I'm gonna have to build a light! :)
Just ordered a shorty S2, I swear those weren't on the site last order I placed just day before yesterday! Now I'll have to wait for that to come in so I can complete this quad. :)
In response to the text I bolded: It’s still 10. A triple of 100-lumen emitters where each emitter puts out 10 less light (90-lumens each) is now putting out 270 lumens. 270/300=0.9, or 90. 30/300=0.1, 10.
While we are sort of on the subject (OK not really), I’ve just been looking at Ledil’s product line and it seems to me that the Sandra series can be cut down to a 35mm triple without the squished-up optics of the Cute and with less depth. (11.1mm vs 15.0mm). Assuming one was inclined to do such a thing. There isn’t an XR or XM version, but I assume that the XP version can be trimmed as described in the OP and/or using the method described by C_k elsewhere.
EDIT: While I’m writing it all down in this post… I speculate that the Cute-3 uses squished 20mm optics while the Sandra-12 uses ~16.6mm optics.
I'm not sure how much is lost, but there has to be a little bit. The issue is that the o-ring lands are there, so unless you can re-cut the groove or do without the o-ring, there isn't much. I've thought about opening it up a bit then using some silicone instead of the o-ring, but haven't tried it yet. I suppose that the easy way to test would be to measure the same driver/LED/cell combination outside the light vs. inside the light.
Wight, from the bottom of the cone to the top of the lens is 13.9mm, each round TIR within the triple is 17.3mm in diameter... near as I can tell with my cheap digital caliper.
Edit: I measured 17.4 across on the mounted one that I've polished the face of. You can see the optics better behind the polished top surface. The second measurements are from an optic that is loose but honeycombed on the surface for the diffusion.