An engineers take:
If you consider a volumetric flow, Q [m^3/s], and a flow area through the media, A [m^2], the filter with the lowest superficial flow velocity, Q/A [m/s] will yield the highest interaction time between the media and the fluid for the bacteria to do their job. This makes sense because it allows the bacteria in/on the media to utilize the nutrients easily.
Since the rate at which the compounds are converted should be independent of local concentration, there is no difference if the compounds are removed by a slow flow or high flow (higher local concentrations will exist at the onset of filtering with slower flow than higher flow if in both cases the media convert at the same rate). I'm not quite sure about this statement. It seems ambiguous & I would appreciate some clarification.
For a slower flow rate, and constant source from waste, the concentration will be higher in the slow fluid because it is being spread out over less fluid. However, a higher Q will yield better mechanical filtering yielding a "cleaner looking tank". Moreover, since the time scale of liquid movement due to filtering rate should be far higher than the concentration diffusion rate (dependent on the square of concentration which is already small before squaring) that the location of the sources should not matter, but again yields a cleaner tank and keeps the media from clogging.
Keeping the media from clogging is more important, since fouling will decrease the effective surface area, Again, I am assuming you are talking about the bio-media & if so, this is a very valuable point. Clogging the bio-media with fine particles & detritus will reduce the amount of surface available for the bacteria to reside in/on.
raising the superficial velocity, thus reducing the interaction time of the fluid with the media and require more "passes" of the fluid through the media to remove the same amount of compounds than a non-fouled media. Just a technical rant...