From the pictures of his exhaust system, it appears that the pipes were supporting the muffler can.  My understanding, is that the entire exhaust will glow red, and the metal will soften (as in almost melt) at high power settings.

The melting point of inconel is over 2500F.  Things will start to glow at about 1000F if the light isn’t too bright.  It is still quite strong at our exhaust temps.  It is worth noting that one of the features of expanding the gas into the large tube is that the temperature drops (Charle’s Law of expanding gases); the temp of the large tube is somewhat below the temp in the header (how much lower depends on too many factors that I don’t know) and depends on backpressure and engine rpm.

Al W. brings up a good point. I had been primarily considering thermal stress, but stress due to vibration from the engine  (aside from exhaust pulses) could be a major factor.  Some high frequency resonance with  engine vibration could do a job on the welds where joe’s broke; or a resonance with the unsupported exit pipe?

I guess I’ll learn some more as I accumulate some hours on my plane.  Now a real challenge is coming up with some kind of damping support . . . .

Al G.