I know for a fact that we can get to at least the boiling point of thallium metal (in Tl spectral lamps), which is 1500C at atmospheric pressure. But even metals like tin and lead have higher boiling points than that, 2600C and 1750C respectively. Zinc has been used in spectral lamps, and you would think that it's boiling point would be higher than tin and lead but it is only 900C. Even gallium has a boiling point of around 2400C. Anything higher than 1600C will soften quartz, anything above 1750C will soften alumina (so lead is probably excluded). Making the metals into iodides makes them boil at lower temperatures, but that's just a MH lamp, not very original. But:

What I would love to see is a potassium SOX lamp (POX?). The boiling point of potassium is a little lower than sodium, and the same two-ply glass would probably be able to work for it. Maybe even having a mixture of potassium and sodium in the lamp, because potassium makes purple and sodium makes yellow and yellow+purple might equal white-ish. Or maybe you can just add a small amount to increase CRI a little. I think this would be very interesting to see.

As to the colors produced by the metals, I have had an okay amount of luck using ChatGPT to generate HEX codes for different metal plasmas, but I doubt that is very accurate. I really don't know how else you would tell, because just looking at a spectrum distribution is not so easy.

As to electromagnetic arc confinement, I would think it would still rely on the coldest spot temperature of the arc tube, but I am unfamiliar with that concept so I don't really know. I would think that for a few seconds you can vaporize almost anything (before the arc tube explodes), but I could be wrong.

My understanding is that not a lot of substances are suitable for arc lamps, and we tried most of them. The only seemingly practical idea that I have come up with has been potassium, but I can't say I have been trying super hard to come up with one. I am sure there are some things that we haven't tried yet. Look into it more, I would like to hear more ideas. I might make them into drawings if they are very unique.

There have been a few already installed around in France and visible in google maps' streetview, i'll send a few links as examples.

Bois-Colombes, France

Eaubonne, France

Eaubonne, France (again)

Mercury and sodium have higher vapor pressure and low vaporizing temperatures. Other metals have lower vapor pressure and higher vaporizing temperatures.

So we have sodium and mercury lamps. Which heat up a vaporize said metals to form plasma arcs.

What is stopping us from upping the heat and vaporizing say... copper, iron, tin, maybe even uranium, gold, lead, Into a plasma arc?

What colors would be produced by such vaporized metals?

I am sure there is alot of fundamental knowledge I am missing. But from my understanding with a thick enough quartz wall and large enough electrodes or electromagnet, it maybe possible to vaporize anything into a plasma state for atleast a short while yes?

I am assuming this has been done somewhere. But I am having trouble finding examples on the web. And finding pictures of said discharges is hard because alot of the search engines also pull up pictures from movies/news articles/market listings for unrelated stuff.

The most important thing in light color/temperature (and light source technology !) is diversity. You may not like 6000K..6500K lights, but the fact they exist and are installed somewhere, make 2700/3000/4000K lighting in other places not be taken for granted, which is an important part of the lighting looking good

I’ve found they only tend to stay white longer as they age?
When new all sizes to me have the same ‘white’ time when running up.

That is correct.  Especially for high pressure sodium. 
Some brands will start up and warm up quicker as well.  Sylvania would warm up fairly quickly versus GE which wasn't bad.  Philips seemed to take a little longer. 

I myself am not happy with the 5000 and 6000k selections.  I much prefer 3000k, 3500k, and 4000k.  I order most of the LED lights I use now from 1000bulbs.com where I can select which colors I want, or get multi-selection LED tubes with those color selections, but I have noticed at Home Depot and Lowes that they are selling color selectable with those 4000, 5000, and 6000.  6000 is flat out stupid.  I know of no environment where it looks better than 4000 or 5000.  3500 is what I consider neutral white and it has become my favorite to use in many buildings, however- there are still some buildings where the color of the carpet and walls will dictate using 3000 warm white or 4000 cool white for best results.  I like having a relaxed inviting color tone like 3500.  4000 works great in building bathrooms and shop areas, and 3000 and 3500 works great in classrooms and offices. 

Bumping this as I am still looking for one. If anyone comes across one please PM me!

An grey Fisher Pierce would probably look decent and was available in the 70s.

There's some from the 90s for pretty cheap:
https://www.ebay.com/itm/305372491904

The oldest one I found was this from the early 80s but it's much more expensive...I personally wouldn't get it.
https://www.ebay.com/itm/276701670178

I've looked for those but I can seem to find one. I mainly going after those tall blue ones that they made around that time such as this one in this eBay listing: https://www.ebay.com/itm/167947718817