Well, post #2 was mine, and was a failed attempt at humor, until more info was brought to light by the OP. Having worked on a number of camera meters, I've see the wide-open metering can differ from the reading obtained by stop-down, among the 2 lens styles. This points to a weak meter cell on the camera with such a big variation from all other instruments used in the comparison tests.
I maintain the same position now as in the corny analogy; that camera 1 and meter 1 were out of agreement by a full stop. At that point the introduction of a second camera would only confound the finding of the culprit as to why camera 1 and meter 1 were a stop apart. So far, not a single piece of gear mentioned ended up close enough in agreement to know where the accuracy was. Chasing down the truly accurate meter is one of the toughest things I've ever done. But it needs to be found in these kinds of tests. One of those camera meters or one of the hand-helds is the truest, and emerges as the standard.