I was warned by nailbender how fragile the dome on my P60 SST-90 drop in was. I dropped it in the garage, and in my haste to blow off the dirt I used a large compressor - and promptly blew the dome off.
I hadn't read this thread, and was ready to write it off as an $80 mistake. I had loved the huge even hot spot with almost no transition to spill. For the heck of it, I unscrewed the reflector and gently cleaned the bits of silicone (remnants of the dome adhesive, I assume) off the emitter. Put it back together, and looking straight down at the reflector was a little surprised to see nothing but yellow emitter, edge to edge all the way to the center.
I have a spare bandsaw blade (circular, 24" diameter) hanging from a nail on the garage wall, this is what I use to compare flashligt hotspot size standing at a concrete expansion line 10' from the wall.
Domeless, the SST-90 hotspot was less than 1/2 the size of the hotspot before the dome came off. And it seemed bright! Noticed a couple rings toward the very outside of the beam, but not bad at all.
Next, my informal throw test - there is a tall palm tree 210' from my back patio, and I use it to compare throw of my lights always compared to a 3D Maglite w/Malkoff XPG drop in, 1.1a at the tail.
Since my SST-90 drop in was only a week old, and I had just done my throw comparison the night before the unintentional dome removal, I had a good memory of how it compared to the 3D Malkoff. I am able to see and count all 18 palm fronds of various sizes and at different angles with the Malkoff. With the new (domed) SST-90 P60 drop in, I counted 12 with ease.
With my D36 SST-50 drop in, (granted, a 3000k tint high CRI with considerably less output than a white SST-50) I am able to count 13 of the fronds clearly. With my P4 modded 3C Maglite, 15 are visible.
With the domeless P60 SST-90, I was surprised to count 15, showing much better throw than the 'undamaged' SST-90!
For overall output, I use an equally informal but pretty repeatable test - flashlight sitting upright on my nightstand, camera tripod mounted in the middle of bedroom aimed at a large picture hanging on an opposite wall. I always use the same camera manual settings including white balance and take a shot of the picture. Only the near wall and ceiling bounce can reach the picture, so it should pretty fairly measure overall output.
The picture shot using the domeless SST-90 appears identical to the domed one I had taken the night before. The Malkoff 3D is dim by comparison in this bounce test, btw, but it is so tightly focused it out throws anything I have short of my trusty old Costco HID.
So, the domeless SST-90 (at least with the reflector nailbender uses) throws much better, and seems to have identical overall output.
I've noticed no tint change yet, but only have 30 minutes runtime on it domeless. It is the 5700k tint (GJ?)
I was ready to trash it after blowing the dome off, but found that I've just turned a flood light into a semi-thrower with a perfectly nice beam, so I'll be keeping it as is. I've already ordered the 'linger special' (2 XPG's mounted side-by-side) to take its place as my go-to flood light drop-in.
So much for my old beliefs that the dome was critical for light being properly directed to the reflector, and that 30% - 80% overall output is lost without a dome (don't remember where I had read that.) Nice to discover that what I've found is not all that different from others here, and it my drop in is not ruined afterall - just changed in beam characteristics.
John F