An assembly house apparently substituted some 1210 zero-ohm resistors with another mfr's part. They're passing >10A, so part characteristics matter.
The substitute parts measure 12.5 milliohms, dissipating an extra 1.3W in an enclosed unit that doesn't appreciate that.
The Vishay parts spec'd list 4 milliohms max, but actually measure 2.5 milliohms. Much better.
Just for fun I measured the drop across the solder fillets too--only 1mV out of 126mV total. Insignificant.
I figure they're both out of tolerance though, one by infinity percent, and the other by 5*infinity percent.
Someone put those in for