I'd guess that it's a lot easier to maintain the whole build chamber at 500° than to maintain the hotend at 850°, but I haven't tried it.
Felsic lavas (and magmas) which melt at those temperatures do not typically contain a lot of alkali oxides, but they do contain some. See https://en.wikipedia.org/wiki/Calc-alkaline_magma_series#/me... However, ferrous and quasi-ferrous alloys like inconel are among the best choices for alkali corrosion. For example, table 4 in Birgitte Stofferson's dissertation https://orbit.dtu.dk/en/publications/containment-of-molten-n... gives an inconel corrosion rate of 1.06 mm per year in molten NaOH at 600°, which happens through oxidation from oxygen dissolved in the melt. Monel 500 corroded only 5.06 mm per year at 700°.
If you were trying to keep a 100μm hotend aperture within a ±10% tolerance, you could start with a 95μm aperture and replace the hotend when the aperture had expanded to 110μm. At 1mm/year those 15μm would be 5 days of printing time, which seems like a usable hotend lifetime. Presumably printing in lava rather than 100% NaOH would extend the lifetime further.
Felsic lavas (and magmas) which melt at those temperatures do not typically contain a lot of alkali oxides, but they do contain some. See https://en.wikipedia.org/wiki/Calc-alkaline_magma_series#/me... However, ferrous and quasi-ferrous alloys like inconel are among the best choices for alkali corrosion. For example, table 4 in Birgitte Stofferson's dissertation https://orbit.dtu.dk/en/publications/containment-of-molten-n... gives an inconel corrosion rate of 1.06 mm per year in molten NaOH at 600°, which happens through oxidation from oxygen dissolved in the melt. Monel 500 corroded only 5.06 mm per year at 700°.
If you were trying to keep a 100μm hotend aperture within a ±10% tolerance, you could start with a 95μm aperture and replace the hotend when the aperture had expanded to 110μm. At 1mm/year those 15μm would be 5 days of printing time, which seems like a usable hotend lifetime. Presumably printing in lava rather than 100% NaOH would extend the lifetime further.