I think we have used just about all of these methods. Depending on how dirty they are they will clean up nicely or not at all. Muffle furnace works well, as does solvent cleaning for organic deposits like tars and char. We can do batches of 10 or more at a time so it can pay off if you have the time to do it.
Cleaning them for use in PCB or PAH analysis should be fairly easy, cleaning them for pesticide work where you can get a lot of break down is more demanding.
When I was running 8270 Semivolatiles analysis I was in the habit of changing liners every morning before starting the next 24 hours of samples, which was usually 20-30 samples in a run. Change liner and clip about 3-4 inches of column each morning usually did the trick without needing to recalibrate each time.