Here's the problem in a nutshell with your position on this-this obsession with reducing blue light over and above everything else. Neither me nor JustAnOldFashionedLEDGuy are advocating specifically for more blue light in the spectrum. Rather, we're advocating for more light at the wavelengths with greater response in the scotopic and mesoptic. Yes, because of the way CCTs are calculated that generally means higher CCTs, but not necessarily high CCTs. There's a difference. 4000K is higher but not high. Nobody is advocating for anything over 5000K regardless, and frankly it seems there's little benefit in most cases going above the low 4000s for roadway lighting or headlights.What are you talking about? Both UMTRI studies showed more blue light producing more glare.
Guess what you can also do? You can also cut down CCT while further cutting down blue. Any sort of improvements of cutting down the blue spike could be applied to lower CCT products to bring their blue light content even lower, e.g. the voilet pump Sunlike.
Ideally, it would be nice if through some combination of phosphors and primary pump wavelengths we could get rid of the unfortunate blue spike present in most LED spectra but we have to live in the real world. There are other concerns like manufacturability and repeatability, as well as efficiency. However, the fact remains you need some of that higher frequency light (i.e. shorter wavelengths) to stimulate the rods which become increasingly dominant at lower light levels. You go with a higher CRI spectrum, you cut down on the blue spike, which is desirable (no argument there). But then you're suggesting cutting CCT to reduce blue further. Fine, but this also cuts down on the needed non-blue wavelengths in the mesoptic and scotopic regions, which means at that point you're compromising safety solely to meet an arbitrary goal of having low CCT lighting. And even that goal has questionable merit given the wide range of CCT preferences among different populations. On top of all that, lower CCT LEDs are less efficient. That's an important parameter for nearly all DOTs as most are cost constrained.
Violet-pumped LEDs sound nice in theory but I'm dubious if they'll ever supplant the current blue-pumped ones. The WPE of violet emitters still lags that of blue. They're also more expensive at present. And the Stokes losses converting to white light are always going to be higher. That's just physics. Efficiency is an important parameter in lighting. For most lighting uses reducing the blue content somewhat at the expense of a huge efficiency hit is a nonstarter. The exception might be where you need to mimic a black-body spectrum very closely. We don't need to do that for streetlights or headlights.
You complain me and JustAnOldFashionedLEDGuy aren't linking to research supporting our conclusions. For starters quite a bit of this research is paywalled. I'm not going to link to something that people won't be able to freely access. For another, I've saved loads of stuff on my hard drive (including that UMTRI study you obsess over which I saved back in January 2010) but lots of it is no longer online. There's what is called a working knowledge that I draw on. JustAnOldFashionedLEDGuy has a lot more of that than I do. I've designed LED products but I work mostly on the electronics end, making the drivers. I deal very little with the optical characteristics of those products. As such, I lack the depth of knowledge of JustAnOldFashionedLEDGuy in that area, although I'm sure my knowledge of electronics trumps his. But as a hobby I've had a desire to learn about this topic, especially in areas which affect me personally, light streetlighting or interior lighting.