Some 20 years ago, I saw spotlights (for theatre use) containing a built-in transformer from 230V to 24V and a huge, E40 based 24V/1000W bulb.
(I can't remember that clearly, it could be only 500W.)

It was a plain incandescent (not halogen - no inner capsule) with a regular globe shape, around 200mm (8") in diameter. The light produced was bright white, the colour temperature higher than 230V/1000W halogen projection bulbs did.

I think the principle was just primitive. The filament (in a shape of very short spiral) was so thick and with such small surface that the tungsten evaporating simply didn't matter.

As I couldn't find anything over the Internet, I wonder if somebody has more information or technical data on that. Thanks.

From that time I remember quite complex voltage stabilizers (one "regular" transformer to 24V for the lamp, one magnetic amplifier used as an "output stage" to regulate the exact voltage and a control electronic using two or three vacuum tubes), according to their manuals intended for high power lamps needing very tight tolerance supply voltage to meet their specified output, yet survive for some reasonable time. And depicted were lamps meeting your description...
So I would guess these were by far not as insensitive as they appear, even when being huge and handling such power, they were very electrically extremely fragile...

From that time I remember quite complex voltage stabilizers (one "regular" transformer to 24V for the lamp, one magnetic amplifier used as an "output stage" to regulate the exact voltage and a control electronic using two or three vacuum tubes), according to their manuals intended for high power lamps needing very tight tolerance supply voltage to meet their specified output, yet survive for some reasonable time. And depicted were lamps meeting your description...

I didn't have this kind of spotlight disassembled, so I don't know if there was an overvoltage protection but the whole unit was dimmed in the same way as 230V spotlights, i.e. it was connected to the dimmable socket (0-230V, thyristor driven).

And as for the lamp, I think it was perfectly resistant to short high-voltage surges. Its filament was a "mass" of tungsten. When turned off abruptly, it faded few seconds out to total darkness.

And as for the lamp, I think it was perfectly resistant to short high-voltage surges. Its filament was a "mass" of tungsten. When turned off abruptly, it faded few seconds out to total darkness.

It is a mass of tungsten, but operated very close to the limit, so even small extra temperature rise could kill it. Indeed, few 100us pulse won't mind, but a second or so overvoltage of just 10% could be killing.

Your setup was probably some of the more modern kind with the thyristor regulation. What I spoke about was a cabinet of about 0.5x0.5x1m, definitely not part of the lantern.

It seems there's a technical maximum of 3200K for Osram studio halogens. Starting from 230V/500W (I = 2.17A) up to a 230V/20kW monster (I = 87A), there is always 3200K. The lifespan specified varies:

230V/500W = 200hrs.
230V/1000W = 750hrs.
230V/2000W = 400hrs.
230V/5000W = 400hrs.
230V/10000W = 350hrs.
230V/20000W = 350hrs.
This is a little strange. They could have typos in their specs.

The melting point of tungsten is 3683.15K. If the rated filament temperature is 3200K at maximum (even if this is a "wire" heated up by 87 Ampers, thus the unevenness in thickness should be relatively minimal), they obviously make the lamp to survive a overvoltage (allowed 10%?) at least for a while.

It is not just about melting point, but about a tendency of local overheating: If a spot reaches higher temperature, it's resistance would be higher than the rest of the filament, so the power density would be larger, so the temperature increase even further.

And the lamps you've listed were rather long life types. There are lamps in use with rated life around 25hours, with filament temperatures up to 3500K These are the types needing the precise stabilization. The uses are mainly the photographic (500W and above) and laboratory (usually halogen capsules with 500W and below). The laboratory ones even do not tolerate the temperature ripple caused by the AC maions, so have to be fed from a dedicated, quite complex, electronic supply unit, by a rectangular LFAC with precisely regulated voltage.