I don't know if anyone dealing with the high-resistance meters/electrometers hangs out here, but I tried multimeters some time ago for another question about the 4339B, and got no response.
I have a 4339B, which was sent to Keysight for calibration. The voltage outputs are in spec, the resistance readings are in spec, but it failed a Keysight calibration on current, when it is reading 0.082 nA low with a test current of 10 nA, but the specification is +/- 0.063 nA. So it's not massively out of spec, being only 0.2% out. It is so close that the resistance measurements are all in spec.
I've attached a copy of the cal certificate.
I've been told this can't be adjusted, but I'm a bit puzzled if there is only one adjustment for the whole of the meter. Given it reads current over many orders of magnitude, I would have expected it to have more than one adjustment.
I actually worked out that if all the currents were scaled by some constant factor, each would have been in spec, but it would be quite marginal. To get the lowest reading value within spec, the highest reading one would be very close to the limit.