But so what?

I for example deeply hate formal suits, and like to wear teenager styled clothes even at work
Call me immature now...

For external light-diffusing material of a LED panel a manufacturer may choose pretty whatever cheapo material he wishes. Any cheap milky plasic sheet.  Of course it can yellow over time even in sunlight.

Well, my guess on S54 is based on the fact that Philips decided to homologate European SDW lamps parameters with regular SON lamps of the same wattage, only dropping the power further as the lamp ages. *May be* but not certain that Philips did the same with their American low-voltage SDW, homologating them to American low-voltage SON. If so, a new lamp will run more or less OK on American low voltage SON gear.

Some caveats:

SON lamps generally require ~1.5kV peak ignition voltage while SDW, at least European ones, are rated to full MH-like 3.5-4.5kV. So SDW lamp may refuse to ignite on a regular SON ballast.

SDWs, having just bare tungsten rods as their electrodes, none of starting aid thin tungsten wire coils as with most HIDs, are having troubles to attach the arc and do glow-to-arc transition for quite long time after being turned on. May be up to 30 seconds for both SDW and SDW-t, even on electronic ballast. So the lamp flickers a lot at start, don't be surprised, this is normal.


Regarding extra chokes. If American SDWs are in fact homologated to low voltage American HPS, 100W S54 will fit directly. 150W S55 will need to add some extra impedance in series to drop the current a bit. Something relatively low, I don't know what you have on hand, may be even a secondary winding of some high-watt CWA or HX transformer with a primary shorted.
 
It is relatively easy to tell if SDW is running at optimal power by monitoring its color and if you can, an inverted black gap over a sodium resonant orange line in the spectrum. If underdriven, SDW will never run up to its proper color, and will look like a regular SON color with relatively narrow gap in the spectrum at orange. If overdriven, the lamp will pass its optimum golden color at 2500K CRI 85 as it is running up and will become dim blueish-whitish with a very large black gap in the orange part of the spectrum.  Turn it off immediately and attempt to reduce the current.

@RRK: I meaning degradation like this: https://www.lighting-gallery.net/gallery/displayimage.php?pos=-223562
@AsXSn said to me that this can't be acrylic because it is UV resistant, and think it is polycarbonate.

I won't say the plastic is degrading fast considering that extreme loading it experiences in the vicinity of semiconductor crystal. As far as I heard it is either silicone or epoxy. Acrylic is popular for external beam-forming elements ('waveguides')



 

I have observed swirling mostly on brand new lamps and used lamps that have been put into storage for a long time.

IMO swirling occurs when the point of contact between the discharge and the filament is too small. This can be caused when the emitter is still in its initial state and the electrode is prone to building hot spots.
It also can occur when the filament is damaged or the emitter is worn out - this can often be observed in flyback-driven lamps (simple battery inverter style). They tend to swirl when the go EOL...

A small point of contact can also be caused when the lamp is mercury starved and the argon is the main carrier of the discharge current.

Swirling is certainly caused by impurities in the tube volume and is not related to cathodes behavior. It happens even in cold cathode lamps and even with non-activated electrodes like in old Soviet sign tubes.

Note that however because the tube is a closed volume at near vacuum, cathode problems can release impurities and this can start swirling.

@RRK: What is the plastic that is common for making LED panels that degrades fast?

AFAIK, so far no one bothered to measure actual current set by this ballast or attempted to reverse engineer the schematics...

Polycarbonate goes yellow over time. AFAIK this can be reduced by some stabilizing additives like antioxidants *in the plastic volume*. Probably not 100% effective and is a subject of manufacturer know-how.

Acrylic does not yellow, but is fragile and has relatively low melting point.

Note good ol' mineral glass can withstand high temperature, is not particularly expensive and of course does not yellow at all.

Aww, that's a shame, I enjoyed looking at your drawings. I hope that you can find an alternative.