I made a post here comparing some of the physical and electrical differences of residential and commercial fluorescent ballasts:
https://www.lighting-gallery.net/gallery/displayimage.php?pos=-264949

What other differences are there? The residential ballast says nothing about thermal protection on it so I am thinking that might not be present there. I hear that 48" F25T12s can only be run on residential ballasts, is that because they normally underrun F40T12 lamps?

The commercial ballasts generally have copper/thicker aluminum windings so they don't heat up as much.
Most of them are high power factor, because commercial/industrial customers get billed for power factor as well.

The reason a F25T12 cannot be run on a commercial ballast has something to do with the capacitor. It is the same reason
you should not run an F34T12 on a commercial ballast. It will make it overheat (or at least run hotter). If you run
an F25T12 on a commercial ballast it will also run at full 40W power.

Here's a discussion about the F25T12 lamps: https://www.lighting-gallery.net/gallery/displayimage.php?album=7829&pos=14&pid=253074

Yes, I did notice that the aluminum windings on the one I disassembled looked pretty thick. I don't think iot did much though because the top of the light fixture when running for a while was still too hot to touch.

So essentially both ballasts run lamps at similar power, but since the commercial version is CWA (more constant-current than HX) it will overdrive energy saving lamps. Very cool.

I assume you are referring to the 4ft F25T12. The reason that lamp can’t be used on commercial ballasts has nothing to do with the capacitor or the ballast itself. The “F25T12” is nothing more than a regular F40T12 with weaker electrodes that can’t handle as much power. It’s a F40 that is designed to be underpowered. This renders it incompatible with commercial ballasts which are usually “normal ballast factor” (operate lamps at their rated power). The F25T12 would operate at 40W in this configuration, causing early failure. Residential ballasts are usually “low ballast factor,” and typically operate F40T12 lamps at 25W or less, meaning the F25T12 gets its designed power and will last its rated life.

F40T12 34W lamps will cause any ballast to run warmer, residential or commercial. In fact, earlier residential ballasts were known to get unbelievably hot with energy-saving lamps. I believe this is because the 34W lamps cause actually draw more power on LPF ballasts than 40W lamps do, which causes the ballast to operate well above its rated power and temperature. Later models like the Advance HB-234-TP were designed to handle 34W lamps better, but you should never run a 34W F40 on any residential ballast that does not say it is compatible with one. Same goes for HPF  commercial ballasts. Many units from before the 1980s were not designed for energy-saving lamps and would overheat with them, so just check your ballast before installing them. Newer ones are usually OK with them.

@rapidstart_12
Yes, I did look more into this. 4ft F25T12s have roughly the same voltage drop as F40T12s (~100V), so CWA characteristics almost certainly don't really play a role here.

Interesting to hear that residential ballasts underpower lamps, I would hope they would make up for that with extra cathode heating to ensure long life, but I doubt it. That definitely explains why they made the 25W version.

Since F34T12 lamps actually do have a lower voltage drop (~80V as opposed to ~100V), that causes their current to go up, loading the ballast more. But just like you said this increased ballast load is usually more noticeable in residential ballasts because the current increase is more pronounced with an HX (residential ballast) circuit than a CWA (commercial ballast) one, so more likely to cook a residential ballast than a commercial one, although of course both have cooking potential.

I did not know that special energy saving rated ballasts existed, I would assume that the 34W tubes were designed to be able to work safely in any ballast, but I suppose that was not the case. Interesting.

All seems so wild to me on the other side of the world....

Why would effort be made to produce dim, inefficient, big, and requiring expensive ballasts F25T12 for home use, and then put them in twin fittings anyway ?

when a single F40 does as good job ?

Or F20T12 at its normal power, half the size, with just a choke ballast, would put out the same light as one F25...

(And if it has a few Lumens less, then overdrive it by a couple Watts to get those Lumens up....)

@Ash - Don’t worry, our residential single-lamp fixtures aren’t free from underpowered ballasts either. In fact, I think that’s where they started. The whole thing really is one of the biggest scams in lighting history. If the ballast underdrives the lamp by a few watts, fine. But under 60% power is ridiculous. And the companies still advertise the full lumen output on the lamps and not even mention that the fixtures are underpowered, so customers would buy these powerful-looking fluorescent fixtures expecting to get commercial brightness but would instead be met with regret and disappointment.

All seems so wild to me on the other side of the world....

Why would effort be made to produce dim, inefficient, big, and requiring expensive ballasts F25T12 for home use, and then put them in twin fittings anyway ?

when a single F40 does as good job ?

Or F20T12 at its normal power, half the size, with just a choke ballast, would put out the same light as one F25...

(And if it has a few Lumens less, then overdrive it by a couple Watts to get those Lumens up....)

The brightness of full power operation would be way too much for the domestic use, because how close these fixtures are.
Yes, it would be possible to use a diffuser with a single F40T12 tube, but that is either very expensive, prone to getting dirty or inefficient (wastes, absorbes, a lot of the light). Mainly in the past there were no suitable materials making a good, efficient long lasting diffuser for low enough price. When using more lamps and reducing the power density you get away with bare bulbs, so no cost for the diffuser, no light lisses in it, nothing to collect dust that is then blocking even more light, nothing to yellow down,...
Plus shorter tube costs more to make than the mainstream size made in huge quantities (because it needs retooling, not just replace the filament feed and the etch stamp and keep the rest of the machinery the same as with the mainstream F40T12), so also a cost benefit.
Plus two lamp ballasts cost about the same as a single lamp full power ballast, so no saving there.

Using two physically large lamps and feeding them by reduced power was just a cheaper way to get what was needed.

Being T12 the lamp is fairly diffused already, its intensity (when looking at the lamp) is not as high as T8-T5. Diffusers made of PMMA sheet formed to the wanted shape with vacuum were used in luminaires already in the 70s if not 60s

The F20T12 was already made, its not like the F25T12 could eliminate all demand to F20T12 ever

And a choke (of the size of a PL choke for us) would definitely cost less than autotransformer ballast

But the short F20 would have all the light concentrated into the short tube, so too bright.
The need is for lower intensity, larger surface light.
Yes, PMMA were used quite long, but if it needs to diffuse the light, it needs either some difraction pattern (the ridges then tend to hold the dust), or have some "milky paint", which tends to absorb a lot of light.
But if you look at common pricing, the F40T12 cost was about the same or frequently lower than the F20T12, the only reason being the F40 was manufactured and sold at way higher production volumes.
Originally the reduced intensity "residential" ballasts were designed to run the common F40 tubes, just at about 25W. Then dedicated 25W tubes appeared, with just thinner filament, to reduce the cathode heating power consumption when the cathode does not need to be that beefy for the reduced power. It is true the 25W version become a lower volume odd wattage tube (although sharing way more production tooling setup with the main highrunner than the F20T12), but the standard of using the lower power for residential applications was already set at the time the dedicated 25W version tube was introduced..

F20T12 have about the same intensity as F40T12, a little less (1050lm 2ft F20T12 vs 2500lm 4ft F40T12 for 765)

I am trying to understand why then bare F40T12 was cosidered too high intensity in the US but ok in Europe

And why over the years bare incandescent lamps, T8, T5HO, and now even bare LED chips without any diffuser (in decorative luminaires for living room no less) became considered ok. What changed ?

The F40 at full power has about the same intensity as the F20, but that is too much for some applications.
The lower intensity of the F40 operated at 25W was used not as a replacement for GLS, but more like equivalent of the linear incandescents used in Europe.
The refractors were used with incandescents, with classic bulb format there was no other choice, but they always had those disadvantages.
Using bare tubes operated at lower brightness meant just a cheaper, simpler and even more efficient way of getting similar result and getting rid of those disadvantages.

In Europe in a "progressive energy efficiency" legislation was formed (somewhere in the 60's till 70's) banning all gear that operated the lamps at lower than rated lumen output, effectively banning an equivalent approach in Europe. Not that European market wouldn't want similar solution, but it was the bureaucratic legislation written by activists what stopped that.
So way less efficient linear incandescents were used instead. It made money for the lamp makers (the lamps were very expensive; so the lamp makers were happy) and consumed way more power (nobody cared), but did not violate the standards (that is what the bureaucrats cared). The modern "energy efficiency standard" laws are nothing new in their stupidity...

Then what changed since then which made bare T8, T5HO and now LEDs acceptable ?

When the further or higher the fixture is, or the higher the general illumination level is, the higher intensity gets acceptable. This has never changed. And installations using indirect lights are not "bare tubes" from this perspective (when the tubes are not in the direct view).
But there were many installations like bare tube lights on a wall behind a bench at the eye level or so.
There the bare tubes are too bright. But if you put them above, there is no problem.
Of course people adopt to what became available: If low brightness tube ballast were available, placing two bare tubes directly on the wall became the easiest and cheapest thing, even when they are in the direct line of sight.
When not (like for the T5's), a fixture with a single tube and a refractor to spread the light over larger area was chosen. The higher price and the need to clean it from time to time accepted as a fact of life, when there was no other choice.
Or it was not installed on the wall, but onto some brackets or hung higher above the desk, so it was not at the eye level. Again a more complex and expensive setup, but if you have no other choice...

By the way in these applications this is the reason the T5's were in fact not more efficient than even T12: T8 or mainly T5 required a refractor, which absorbs a lot of light, so you need higher lumen light source to compensate for the losses. With a T12 tube, which could be run bare, the less losses means less lumens are sufficient, so even with the lower raw efficacy of the lamp the overall consumption becomes comparable.

LEDs can not work without a cover for electrical safety reason, but there is quite wide range of intensity to select from, so you use the lantern that suits the need. But the advantage of LEDs is, their lifetime allows to make the refractors sealed, so rather immune towards dirt accumulation on the rough side of the refractor (where the dust tend to stick and becomes difficult to clean out). Plus the LEDs do not expose the plastic to the UV, so there is wider selection of or less restriction to materials for it, so it could become cheaper.

All luminaires compared are ceiling installation, and all really have no diffuser or any type

The Fluorescent ones with bare tubes (36W T8 or 54W T5) are installed flat on the ceiling, at about 2.5m height

The LED ones with bare visible individual LED dies are often in the form of dangling chandeliers, those are often installed in newer houses with higher ceiling, but the actual light emitters are hanging at about the same 2.5m

Many of the LED ones are only counting on isolated LED drivers for the safety, some are questionable at that too (output Voc exceeds SELV limits, bad clearance distances on the PCB or inside the transformer insulation are common)