The error of a multimeter usually is much higher than +/- the lowest digit. So changes between measurements could stem from the multimeter and not from the DUT.Yeah, I suspect that when measure at different time, the number might be different by 0.001V, due to error range of multi meter. But, I expect the number to stay stable when doing measurements.
However, as you wrote ...
... that the X16 is more stable than the E30 hints that it's not the multimeter which is wrong.It was interesting to note that a E30 could not stay stable for a while when cold boot. X16 was very stable (powered on for a full day before measurements).