@med, I still don't quite get this addiction to getting meters to agree with each other spot on to the last dp and leaving things on for that duration to spot drift and tear into it if any is spotted. We all accept that until a few years ago, the military of both our great countries used Fluke 25 and 27's as their meters of choice, I myself have a number of these ex NATO meters and they were only 3.5 digits. So, to my mind it is far more sensible to only really worry about meters being accurate to the third decimal place.
In the real world, i.e., in the military, there is no such thing as leaving meters running for hours at a time before using them on equipment. Its a case of, having a problem, lets get the meters out, switching on and checking the faulty equipment without any delay so that it can be brought back into full readiness condition in case it is called upon in an emergency.
Yes I have 8 bench meters and I have a couple of calibration tools and the one most used is like you, a AD784 but I use it to give me reassurance that a meter has not suddenly gone rogue giving me way off readings. Last night I connected, all 8 bench meters (4.5 to 6.5 digits) together to my AD584 and switched on and once they had all done their self tests, all agreed with each other up to the 2nd dp and the 3rd dp within 1mV of each other and that in the real world is perfectly acceptable. I refuse to allow myself to be concerned beyond that because that way lays madness and an ever deepening rabbit hole and money pit, for no real payback.
To me, those extra digits come into their own if you want to do a quick check to see if something is charging or discharging because they allow you to spot movement up or down far quicker, apart from that, until I see quoted values in a parts list or on / in a service manual / schematic that quote values to that accuracy will I ever worry it.