Here are the new shunt (on the left) and the old shunt next to each other:

Notice the discoloration and warping on the old shunt? Here is the same area from below:

Definite heat damage. It's not clear to me whether there had been a loose connection at some point (generating extra resistance == heat), or whether the 50A rating of this shunt was just exceeded too often, but it is clearly damaged. I'll monitor the new shunt for damage frequently for a while.
Driving with an operational ammeter is extremely useful. For one thing, I can see that my "normal" and "economy" modes indeed are not different - which I had *felt* but not proven. Both pull 50A max. I'll tweak that at some point. It would be useful to have something in between 50A (economy) and 150A+ (performance)...
Addendum: this is very similar to damage that Randy Pollock suffered (although not as extreme):

We have the same shunt, in the same system, and had the same failure. I don't think this is a coincidence.