The count-rates (DLS) will give you a qualitative comparison. In any case magnetite has a strong optical absorption at ca. 600 - 650 nm. UV-Vis spectroscopy to get extinction coefficient (at a well selected wavelength) is the most accurate and hassle-free method for you I guess. It is important for both these techniques, to have a stable dispersion both before and after recovery.
For scattering measurements the Rayleigh approximation tells us that Ι α d6 to the power 6.... so the counts of the DLS equipment are related to that... So it is extremely sensitive towards small size variations. So if you are sure your solids variation is the only variation... you may use it as a qualitative comparison.. else not.
Why not do a very simple gravimetric determination? Does the nature of your sample give a problem to do so?