I bought this meter a few years ago, but I consider the 6.5 digit function to be rather worthless. The problem is that the meter's input impedance is only 10M! This means that when measuring anything but a power supply, the meter loading will turn the last digit or two into trash.
But it gives you the resolution to see changes.
Yes, that's true and that could be very useful. But due to the low input impedance and depending on what it's measuring, it might have only 5.5 digits or less of
accuracy.
I measured the output voltage of a Fluke 731B DC Reference Standard with a Fluke 8375A DMM (Nixies!!). Input Impedance is spec'ed as 1G for the 1V scale. The reading was 1V00013. I bridged the PM2534 across the connection and the reading dropped to 1V00004, a change of 90 uV. That's due to the ~900R output impedance of the 731B. So, in this case, the effect was noticeable at the LSD of a 5.5 digit* measurement. As the source impedance of your device under test increases, the effect will get worse.
Ed
* For the OCD members of the audience, the 8375A can't actually be described as 5.5 digits because it counts to 119999 instead of 199999. Still, not too shabby performance for a meter from the early '70's!