The thing is, if you are a paying customer for the weather service, you will get more reliable access than someone trying to look at it for free, mainly when the server load approaches its limits...

The problem also is, the only way to eliminate flicker is by using rather large (for the power) electrolytic capacitors and these are rather problematic components mainly from reliability perspective. Not that it wouldn't be possible to make them reliable enough, but it costs money and mainly building space within the ballast cavity (which uses to be extremely small with modern LED products), so much it forces the other components to operate in not that favorable conditions, increasing the cost there (to compensate with "better" components) as well.
Skipping this capacitor allows you to achieve significantly better reliability for the same budget, power and efficacy.

Why the cheepeese are skipping it is clear - as it allows to cut cost significantly even without giving up the reliability of the rest, along with compromising on the quality of the rest it really saves a lot.
But with "quality brands" I see the market rather went into eliminating flicker with mediocre reliability, I'm missing the uncompromised reliability with tolerating some flicker - as to me many applications don't benefit that much from true flicker free operation, but do benefit from the reliability.

"getter" is a chemical intended to absorb atoms and molecules contaminating fill of sealed structures.
On incandescents a red phosphorus was used to absorb any remaining traces of moisture from the lamp internal atmosphere.

A barium is used to absorb mainly oxygen mainly in vacuum devices - in vacuum tubes it use to have the form of a coating on the glass surface, applied by evaporating the pellet containing Ba after the tube was evacuated (either heating the cup with Ba pellet with laser, or earlier filling thin tube by the BA, welding this tube into a loop and then firing it by an induction heater so thethin tube bursts releasing the Ba) and letting it condense on the tube glass wall. During this it reacts with most of the remaining air in the tube, but then the coating on the glass was absorbing the gasses that got released from the internal materials over the lifetime, so helps to maintain the good vacuum there. It became the characteristic mirror-like finish on the top of the tubes. You may find the same form also in some Na lamp outer bulbs - usually the mirory like patches around the bulb base.

On discharge outer bulbs the vacuum does not have to be that good, but the "mirror like" form was unwanted (as it obstructs light, or even reacts with the IR reflector coating on SOX), it is just put into a cup welded on some stem wire and let it act from there.

Eric, even more interesting when such anomalies as yours begin to turn up!  Its seems there are no hard rules when it comes to documenting lamp designs over time.  Engineers are always playing with ideas for improvements (which sometimes accidentally turn out to be a backward step during the later phases of lifetest, like when they prematurely introduced full-coated bulbs).  So it is not unexpected that they may have made some runs according to the earlier fully clear-dome design to check whether that made any real improvement, and we end up with lamps like yours whose date seems not correct at first sight.

Today I got a little more done... changed the oil in my sport car around 2am in the morning, (took five hours, since the oil filter was stuck.)
and... bought a GoPro Hero 10. First GoPro I've ever bought so far. I'm planning on doing what's called "Motovlogging" with this GoPro... vlogging while
I'm on my motorcycle. Major event: this is something I've wanted to do since 2009, before I knew about Lighting Gallery. For me, this is a big deal.
However... the GoPro features are endless. Insanely complexed... so I'll be trying it out once I have more time. I also need an SD Card, and a microphone
to install into my helmet, etc.
I've also talked around with a few friends to see if they want to go to "Cars and Coffee" in Dayton Ohio on May 25th, since I'll be off work that weekend.

Cost being the most decisive factor, I guess. Now, cost aside, what would prevent say a LED edge lit panel to have a rectifier to run it effectively on DC ?

Fair enough. I used some 1/2" 1/16" thick aluminum bar stock to make the bracket for my Bantam. Drilled and tapped holes for mounting and then used a vise and a small bending die set to make the bends.

But I kind of want a somewhat official one with a real G12 base and custom bracket, kind of like the Chicago spec M250s in CMH (but those have a different base). Besides, I could use the mogul socket and the old ballast to convert another light to HPS.

I noticed that too. Even with LED lamps, the cheap ones tended to cheap out and have noticeable flicker while the more expensive ones tend to have additonal components in the driver that at least reduces flicker.

I observed something interesting this week...

Way back in the late 1930's, when commercially viable discharge lighting came on the market, the magnetic ballast produced flicker at twice the mains frequency. Over the years, different mechanisms were used to mitigate the effect (lead-lag, distribution over three phases) until the electronic ballasts came on the market in the later years of last century. The high frequency made for virtually flicker free operation.

Now, cue in the advent of LED lights, the cheap hardware store ones have very annoying visible flicker. This week, I paid a visit to the electrical supplier. All their fluorescent troffers have been switched over to LED edge lit panels. I walked in and there was no apparent flicker. When I was waiting for service at the counter, I spotted one of those retro-style fans, that have metal blades and a metal enclosure. It was running and the blades had that slow movement appearance that is usually a tell-tale sign of flicker. I was kind of surprised that the LED panels produced flicker akin to magnetically ballasted fluorescent.