Which is actually?

Blue seems to be mixed with green, is this intentional?
()

I'm setting color like this:
nuclearReactorReactor.light.color = {b = 1, g = 0, r = 0}
so it should be pure blue. I think it was working in 0.16 so maybe they have changed something in way color is handled.

I've found two lines in the source of the original reactor which may be connected to the problem:
-- use_fuel_glow_color = false, -- should use glow color from fuel item prototype as light color and tint for working_light_picture
-- default_fuel_glow_color = { 0, 1, 0, 1 } -- color used as working_light_picture tint for fuels that don't have glow color defined
I'll test it when I have some time.

I re enable/disable all modules, it seems to be a problem caused by bob's module...

del

The glow is still mostly white so I would consider it a bug.

There are changes in how the color is handled. I had to search them all out in my nuclear mod. You can see all of the places you need to change if you want to check my code there. I would tell you but I don't remember them all because I made that mod 6 months ago. The main things I remember are that you need to be sure you've handled the fuel's reactor glow color, the default reactor glow colors, the working light color, the light color, and I think there are one or two more. I also found that the colors don't always work as expected. Tried filing a bug report about it but they didn't believe me. :P

If you do look at my code, my colors will be different because I'm making a different assumption than you are. You're using colors from cherenkov radiation that occurs within water from a submerged reactor, whereas I interpret this reactor to be a liquid-metal-cooled reactor, which glows with the red-orange of a liquid metal coolant. But the slew of light prototypes that need to be changed are the same.

I did a few test and it looks like the color works after all. It just doesn't look as blue as I hoped. Maybe because during night everything looks blueish and maybe blue is just similar to white.
For a test, I've changed the tint to red for a test and it is clearly red on the whole area beside a small path around the heat pipe at the top of the reactor which is whiter. I think this spot is partially reserved for the heat glow. The vanilla reactor has this path too.

I would say this case is closed.