https://www.ies.org/fires/circadian-lighting-an-engineers-pe...
- The metrics Circadian Stimulus and Melanopic Lux are flawed
- We don't know enough to produce products yet
I agree with him on point 1, especially about CS. The author, and several others, have demonstrated a huge discontinuity around 3500K. The CS model needs improvement.
Regarding melanopic lux, the author says it is more a problem with our measurement and modeling tools. Most manufacturers won't provide a Spectral Power Distribution, but this is starting to change (we do). There is at least one clinical trial in progress that is testing this metric directly, and several experts think it shows a lot of promise.
With that said, new research on blue vs. yellow/orange opponency is coming out. Basically, we're trying to figure out contribution of the visual system on top of melanopsin. I'm not sure what to make of it yet, but I think we'll find out much more in the next 2 years. We probably need a better metric than melanopic lux.
On point 2, I agree that our knowledge is changing. But it takes a number of years for standards to be developed, much slower than the rate of technology improvement. We are paying attention to all the new research, but we based our design on our best understanding of the research to date.
We know that we can really only attract early adopters at this stage who "get" it, and that most people are going to wait and see until things are more standardized. But we are effectively able to do mini-experiments with our customers, which can lead to more insights/avenues of experimentation.
Case-in-point: f.lux. Research has shown that it is somewhat effective on its own, but combined with dimming, it could be quite effective. That's actually what we're doing with our product—controlling both the spectrum and intensity. And f.lux is able to run experiments with a portion of their users to advance our understanding.
Basically, my opinion is that we have to start somewhere, and we believe we have enough evidence supporting that our solution is "better than nothing." We were able to prove that there is enough of a market for these types of solutions already, even though our knowledge is still evolving.
(edited for a paragraph spacing typo and a 2nd time for a grammar issue)
You may think extremely bright SAD lights is already readily available, but there's anecdotal evidence to suggest they are insufficient
https://www.lesswrong.com/posts/koRZu53LBZEapwww6/could-some...