Well, quality LEDs are expensive not because it is more expensive (anymore, I should add) to make them, but because customers are willing to pay those money for them. They are (relatively speaking) cash cows. And because many key design components are still protected by active patents, there is not that much competition there.
For the older, less light quality designs the patents have expired, so many cheepeese makers jumped that bandwagon and the prices have dropped...

The UV are still (and will always be) expensive, because there is minimum production volume. With any semiconductor technology, this effect is way stronger than with the "classic", because the base production line is many orders of magnitude more expensive to set up, while then it is running relatively for free. So selling 100 millions vs few 1000's means final product price ratio 10000× or more.
With that I think it will be still cheaper to maintain few custom lines for the special discharges.

At the end everything is about money. If something is needed (like the germicidal UV), the cheapest source will supply. If that would be maintaining a discharge line, it wont go away. If something will require carbon arc, someone will be making the rods too. Sure, very niche, but running business.


Edit:
What maz change the game for the germicidal UV lamps is the present China virus mess: Small sterilizers may become a consumer goods for home use, so a large volume item, where the discharges are rather impractical (large size, short life, high large volume production cost).
That may, I think, bring large enough volume to dilute the short wave UV LEDs production line cost, so the unit price may go significantly down. Plus these small units are the best fit for LED (mechanical robustness,...) once the LED becomes cost viable alternative.

What do you think? Those LED thinks are just to laught! 

A video from Big clive: https://www.youtube.com/watch?v=_EVw_AhPJ7E&feature=youtu.be

*Note that there actually real uvc led, but they are made of metal and they are very expensive.

But, the part that is not that funny is the fact that pleople buys those bulbs believing that actually those are germicidal, when those are just blue LED or blacklight.

I think that LEDs can't replace Germicidal lamps.LEDs don't oroduce ozone,and wavelenght is specific in these tubes.

I think that LEDs can't replace Germicidal lamps.LEDs don't oroduce ozone,and wavelenght is specific in these tubes.

Technically they can. You may make LEDs with the same wavelengths as generated by the Hg discharge.
Even the opposite:
With the discharge you are stuck with whatever lines that particular element generates, maximum you may do is to filter out the unwanted lines. If some wavelength is not present in the discharge spectrum, you can just never get it, so have to live with the consequences in some other manner (e.g. deal with the ozone generated by the germicidal lamps).
With LEDs you may simply tune the wavelength to whatever you want, so it is way easier to e.g. tune it so it is still effective as germicidal lamp, yet produce less ozone (so less ozone for the same effect). Or you may tune it so it does produce just the amount of ozone you need for the task. You really have the freedom there.

Of course, one LED could only produce one line, so to get the whole mixture, you need multiple LEDs, one for for each line.
In fact for a commercial product, the LEDs would have to be mixed so it radiate, along the main UV, the same signature in visible as well - for safety reason, to make the lamp visible the same way as people are used to with the discharges).

The only thing preventing LED use for such short wave UV  as the germicidal lamps (at least for now) is the cost of the short wave UV LED emitters. But only because the market for such products is rather small (only professional use so far), so not much benefit from the cheap mass production the semiconductor technology offers given its set up cost (in millions for just interrupting the production of an existing line and retooling it for that batch). If the germicidal applications reach consumer mass market (so that batch of the UV LEDs would become large enough), the picture may becomes quite different.

Today these short wave LEDs are really useful only in application that really need just one specific wavelength with highly focused beam, paarmeters not attainable with the discharge, so willing to pay the associated cost (like the UV resin curing 3-D printing processes,...), just because the only alternative for them (the lasers) is even more expensive.

Technically they can. You may make LEDs with the same wavelengths as generated by the Hg discharge.
Even the opposite:
With the discharge you are stuck with whatever lines that particular element generates, maximum you may do is to filter out the unwanted lines. If some wavelength is not present in the discharge spectrum, you can just never get it, so have to live with the consequences in some other manner (e.g. deal with the ozone generated by the germicidal lamps).
With LEDs you may simply tune the wavelength to whatever you want, so it is way easier to e.g. tune it so it is still effective as germicidal lamp, yet produce less ozone (so less ozone for the same effect). Or you may tune it so it does produce just the amount of ozone you need for the task. You really have the freedom there.

Of course, one LED could only produce one line, so to get the whole mixture, you need multiple LEDs, one for for each line.
In fact for a commercial product, the LEDs would have to be mixed so it radiate, along the main UV, the same signature in visible as well - for safety reason, to make the lamp visible the same way as people are used to with the discharges).

The only thing preventing LED use for such short wave UV  as the germicidal lamps (at least for now) is the cost of the short wave UV LED emitters. But only because the market for such products is rather small (only professional use so far), so not much benefit from the cheap mass production the semiconductor technology offers given its set up cost (in millions for just interrupting the production of an existing line and retooling it for that batch). If the germicidal applications reach consumer mass market (so that batch of the UV LEDs would become large enough), the picture may becomes quite different.

Today these short wave LEDs are really useful only in application that really need just one specific wavelength with highly focused beam, paarmeters not attainable with the discharge, so willing to pay the associated cost (like the UV resin curing 3-D printing processes,...), just because the only alternative for them (the lasers) is even more expensive.

Sounds possible.Germicidal lamps are only used in hospitals and science.

Sounds possible.Germicidal lamps are only used in hospitals and science.

And water purification equipment such as pumps for fish ponds. Also EPROM erasing, the Philips TUV6 lamp I have came out of an EPROM eraser.

And water purification equipment such as pumps for fish ponds. Also EPROM erasing, the Philips TUV6 lamp I have came out of an EPROM eraser.

EPROM eraser is used to reset chips ?Much better than Chlorine tablets(Cl).

Sounds possible.Germicidal lamps are only used in hospitals and science.

I've also seen those in tanning cabines.

UVC..? I hope not 

UVC..? I hope not  

I Mean they only (I think) use UVA and UVB.

Edit:

Will be weird to in 20 years more go to a tanning cabin ad see a bunch of LED.

I've never been a fan of artificial tanning. If your skin was meant to be a particular colour, it would be already. Too many people get skin cancer already 

I've never been a fan of artificial tanning. If your skin was meant to be a particular colour, it would be already. Too many people get skin cancer already 

Artificial tanning is a big NO SENSE when taking your head over your window or just going to sleep in your patio at middle day will be ok and free! I'm shure that I don't need any tannig even natural. 

Aside I'm quite lucky that I have a decent size patio where I can make my homenade high masts.