Very interesting to see that most of them have pretty strong infrared emissions. I always wondered if a mixture of multiple alkali metals could be used in an SOX style lamp to make a higher CRI version of LPS, looks like cesium and sodium would be the best bet for that, maybe a little rubidium for some good blue-purple. I am sure efficacy would go down the drain through

Although this would work in principle, the problem with mixing alkali metals is amalgamation which would have a negative impact on the vapor pressure vs temperature characteristics of the metallic fill of the lamp. A well known example of that is the sodium-potassium eutectic mixture (22/78 Na/K ratio) which is liquid at room temperature but has a lower metal vapor pressure than pure sodium. So, to make this work, the burner's operating temperature should be higher than that of a SOX lamp, which would certainly result in an accelerated glass corrosion from the alkalis (vapor and liquid).

Oof, I know how much of a challenge it was to get sodium to vaporize efficiently, making that even harder certainly wouldn't be worth it. I was not aware of this vapor pressure characteristic. Maybe a high pressure approach would work better, but at that point metal halide is probably better.

Yes, the metal halide approach is way easier. Interestingly there was some research done on mixed metal vapor lamps using HPS burners (those can withstand much higher temperatures than glass/quartz ones). GE looked at the possibility of adding metallic elements such as cadmium or thallium to standard HPS lamps in order to improve their light color. Sylvania investigated lithium-cadmium fills for horticultural lighting. However, none of those works were successful for the reason I mentioned earlier. So, yes, it's metal halides all the way if you want to combine elements in order to produce lights of improved color (and efficacy).