That the self-test circuitry, itself, is the cause of this error is a very insightful response.
SIDE NOTE: For those out there that would like to understand this bug better, the 1271 multimeter runs hundreds of internal tests, and this meter passes all, except 2433. And this is actually NOT a measurement, but a digital calculation of the offset from other measurements that, themselves, all pass; it's a calculation using the output of the RMS converter from the application of +1 and -1 volts recorded during previous tests. The offsets from these tests are averaged and then compared to the Full Scale values; the result must be within +/- 100ppm. The measurements that the digital calculation uses (from Test Path P063 and P064; these I can inspect) are well UNDER the max 100ppm limit. So, that's the dilemma: The apparent source measurements pass...the values are within limits...but the digital calculation of the offset from Full Scale does not pass. It's kind of like saying: A=1 and B=1, yet A+B does not equal 2!
BACK TO THE MAIN ISSUE: There is no way for me to determine what the 2433 digital calculation is actually measuring or calculating (other than relying on the printed Datron documentation for the specific measurement) without single stepping through the ROM code and reading what is being retrieved and stored in memory (not possible, for me, at least). So, to help eliminate the possibility of self-test errors being the cause of this problem, I have installed the RMS board in the second 1271 system I'm rebuilding. I get the same error (and only that error), so I can (maybe) eliminate everything BUT The RMS board.
If one of the test-path relays on the RMS board was not working correctly (which MIGHT cause this type of error), it might corrupt the overall AC readings from the RMS board, which is not the case; AC measurements made with the 1271 compared to my calibrator output are actually OK on all ranges. So the error is likely not in the "meat" of the RMS board (the PreAmp, Rectifier, and RMS Converter). Which means that it STILL COULD BE a fault in the self-test circuitry on the RMS board, which I've examined very closely and tested. But I probably should go back and look at that, again.
Thanks for the great suggestion.
And thanks, again, for everyone that has provided feedback. All ideas are welcome!