Yes -- I think it's the 801 versus the 801b.
I've only had my 242D for a couple of weeks; it has the 801b. (I got it because resistors are my hobby, not because I expect to impress anyone!)
I'm seeing about +/-200nV total noise; I'm not sure where it's coming from, but everything is still pretty workable at that level.
A quick check of the TC of the system, including 1k from the 925, was about 2ppm.
I like that I don't have to put all the eggs in the 925 basket, as good as I think it's going to be. The substitution method removes its calibration from the results, and you measure only the small ratio between your standard and unknown, right out in the open where you can see them. I trust the deviation dial for that small ratio because it has such a nice wire-wound sound and feel. I think the 0.01x, 0.10x, 10x, and 100x multiplier settings are going to need some tweaking -- they seem to be within a few ppm, but I think they could be better.
Fun. Big. Tedious. Somewhat, but not excessively, needy; most of what needs doing can be done with Hamon buildups.