Yea, I think the 3.0 umol/j is most obiviously theoretical. (although, I was actually thinking 2.6 or 2.7 umol/j) But maybe your right. I suspect because the IR and UV in these lights, making it more difficult to easily measure (but an expensive/integral component, takes up resources) So there likely accounting for the unmeasurable spectrum that's not going to show on a standard light meter, probably only able to measure on a mass spectrometry, which is hella expensive.
I've been studying this tech for awhile now, since around 2010, (I've also built my own computers for the last 20 years, watercooling and all) and I've done tons of research. The UV/IR is necessary from my perspective for overall enhanced oil production during the flower.
I've taken 2 of my lights apart the last few days to maintenance them, clean them up, and check them out, (clean the fans and heatsyncs, remove dust from the driver units) then back into service, I've got some avionics (corrosion x) that's really good for keeping them in top running condition. I've also thought about using conformal silicon to enhance there "waterproofing" But I've not done that yet. (I"ve done it to my drones) I'm kinda an electronics geek as well as the cannabis thing. I want my stuff to last, so maintenance is a must. I've got good soldering and electrical troubleshooting skills, so that's not a problem if something happens at some point..
LED's have more maintenance issues than HPS units, especially if any fans are involved for active cooling. That's why I like the newer units like yours, but there not without potential problems either. Ever allowing these lights to heat up, is a really, really bad thing, and diminishes overall efficiency and lifespan, just as you say.
I'll probably open this next light tomorrow, maybe I'll take more pictures of what I actually do inside.