From my experience, the primary problem with metal seals is thermal cycling rather than temperature alone. If you think about thermal expansion, the seal gets better with temperature as everything expands. Thermal cycling can lead to small leaks but they are typically on the 1E-7 or 1E-8 Torr scale.
Depending upon the seal locations i.e. system design, the seals can be protected from shock and plasma. We seldom had problems with metal seals at even modest repetition rate (~5 Hz) that led to ~100 C chamber temps. At higher temp (>400C) you need to look for special seals as the common sealing materials (copper, organics) no longer work. I think folks tend to favor soft iron but that could introduce new problems as the material is magnetic. Organic seals tend to be more resistant to thermal cycling (<100C) and movement.
If you want to put layers of seals you have another problem, trapped volume. If you use multiple levels of seals typically two layers with organic seals you can achieve a higher base vacuum when you pump the volume between the seals to ~1E-3 Torr. The leak rate across the inner seal is reduced by ~1E5 so you can achieve better base vacuum such as ~1E-9 Torr instead of ~1E-7 Torr. For a PF device, you can get away with 1E-6 to 1E-7 Torr base vacuum without any problems. Some folks operate at base pressures as high as 1E-4 Torr. I don’t know of anyone that would use metal seals in multiple layers. Metal seals are good to 1E-11 Torr at the worst. The typical problem with vacuum is virtual leaks. Imperfections in the surface, physical absorption of things like water and gas out of materials are generally the limiting factor in base vacuum. For high vacuum systems, it is common to bake the vacuum chamber at ~100C to motivate the water to leave the surface of the vacuum chamber. I’ve never done it for a PF but I have done it for high vacuum deposition of metals. There are other tricks with chamber material choices like Ti instead of SS304 or coating and SS304 chamber with Ti or Nb.