The F5 and F6 have a difference in RGB metering that would explain the result being slightly off the F4. With the Canon, I don't recall how the metering bias affects the weighting, so tougher to explain that. One thing to consider is that an f number is not always an indication of transmittance; except in some Cine lenses that actually have T numbers. There might be enough difference in the two lenses to account for some of the variation.
I think the only real way to tell would be to bracket a few rolls of transparency film under controlled lighting conditions. I have done this with each new 35mm or medium format camera when I got them, since it is a good way to see if the shutter speeds, aperture settings, or the metering are somewhat close to what is indicated by the numbers. Some cameras consistently overexpose, while others consistently underexpose, though more than a few are close enough to trust the settings as true.
You should also remember that where the meters are placed in the camera body can have a slight effect on the results. A good example is the FE, FE2, FM2, and FM3A, all of which can share focus screens; there is a slight difference in newer screens that affects the meter results. In a similar manner, and accumulation of dust, or even a slightly coated (dirty) optical path would alter the readings. Parts do accumulate debris over years, and electronic components do sometimes where out or fail. It would be tougher to state that the brand of camera has more influence than these factors.
Ciao!
Gordon