> I meant an early revision of the firmware (not the processor board)
> that misidentified the processor because it could work with either.

Firmware and OS identify the CPU by reading and interpreting the content of the PVR register. Unfortunately, 744x and 745x (with same value for x), which only differ regarding the (in)ability to connect external L3 cache, have identical PVR register content so can‘t be distinguished this way. That‘s why normally, such CPU is reported as "744x/745x" (with same value for x) by firmware and OS. Theoretically, 744x and 745x could be distinguished by additionally detecting L3 cache, but this wouldn‘t be completely safe because a 745x without connected L3 cache would be misreported as 744x then.
So I‘m not sure what‘s happening with jalla‘s Pegasos II here, but maybe he is simply misreading "7447/7457" as "7457". Just a guess, as an indication of only "7447" or "7457" would be strange considering the explanation above. Maybe other Pegasos II G4 users can report what their MorphOS bootscreen says.

> I'm fairly certain, like you, that all the Peg2 G4 boards were all 7447 based.

...except the few 7447A-based ones :-)