The 34972A is a nice bit of kit, worth buying over the 34970A if ypu will be working 'in the field' away from the bench.
The USB function allows you to make programs using the Agilent benchlogger software and load onto a USB stick to insert into the rear USB port to allow for stand-alone automated testing.
I built up a test rig recently for testing individual cell voltages on large 52-cell grid-tie batteries out in the field and it works fantastic, the client has me building another setup at the moment.
These instruments are fun things to play with.
On a side note, you can read out the number of relay actuations on the internal relays from the front panel, and the module relays from the front panel or from the PC software. That way you can replace any worn out relays for better test results on second-hand modules.
I found my results were much better and more consistent when I did this, some of the 2nd hand 20-channel modules I picked up had relays were up at 2 million actuations!
If you do a before and after ohms test with wire links to short each channel (so you are effectively testing the relay contacts) you'll see much better consistency across multiple runs when you swap out old relays, which is expected..
Also, the 34972A (and maybe the 34970A, not sure) has a CR2032 coin cell inside. A potential source of cheap units with an easy fix if it's throwing errors due to a flat battery.
Just read the service manual and be careful as there are some screws around the DMM module which if removed can apparently potentially affect the calibration. (But you'll need to remove them if you need to replace any of the three internal relays anyway).
Also, replacement VFD displays are available from the usual online marketplaces. Due to the nature of these instruments, they can be left on for long periods during logging, and the red phosphor in the display tends to dim quicker than the green, so a new VFD can be a nice upgrade if the budget allows.