Before using a bead based approach, i would recommend to verify if you are able to detect small enough differences in volume (1-2%). I guess a high conc. of beads might help.

if not i would use the gravimetric approach.

Will be the same pipette used, or does every labmember has his/ her own?

If it is the latter, you also have to take into account that each pipette has its own error.

I haven't found any literature yet, so if you have any on hand I would be grateful.

We are all using the same calibrated pipettes. So the focus would be on staff qualification.