I did some quick accuracy testing against my 8846A and it looks pretty impressive. I am not a Cal. lab so this is the best I can do.
8846A was set for 100NPLC and digital filtering.
2450 was set for the max 10NPLC and filtering with 10 readings to approximate the 100NPLC of the 8846A
Both units thoroughly warmed up.
2450 source and measure 0.1 volt with remote sensing to the 8846A 2.6µV difference in measured values or 26ppm. Max 8846A error for 1 year is +/-6µV (last 8846A cal. deviation was -0.5µV) Max 2450 error for 1 year spec +/- 212µV
2450 source and measure 1.0 volt with remote sensing to the 8846A 4.9µV difference in measured values or 4.9ppm. Max 8846A error for 1 year is +/-25µV (last 8846A cal. deviation was -1.0µV) Max 2450 error for 1 year spec +/- 420µV
2450 source and measure 10.0 volt with remote sensing to the 8846A 13.7µV difference in measured values or 13.7ppm. Max 8846A error for 1 year is +/-230µV (last 8846A cal. deviation was -0.00000V) Max 2450 error for 1 year spec +/- 2500µV
The 2450 spec on current is way way way better than the 8846A below 0.1A so the only place I can use the 8846A to check is at 400mA where it is 3x more accurate than the 2450
2450 source and measure 400mA to the 8846A 80µA difference in measured values or 201ppm. Max 8846A error for 1 year is +/-180µA Max (last 8846A cal. deviation was +1.0µA) 2450 error for 1 year spec +/- 620µA
I am impressed!