Neil Frank did this, if you move things around and do not do water column dosing,m this method works well and allows you to try out various sediments to compare with(they all leach into the water column, so keep this in mind and how that might make all of the results seem equal and how you might over come this issue).
I've never once had any issues with so called root feeders using the water column.
I often question the term and the notion of "preference".
Plants do not care near as most can tell, but you might not dose routinely, so sediments are good back up in case.
Then you see differences, but you must account for the overall nutrients, not just the ones you measure in the water column, something obvious yet overlooked.
I can add the same amount of NO3 to the sediment and the water column and have the growth relatively the same. Several nice studies showed this and the general consensus among researchers studying plants, they are opportunistic, they will get nutrients wherever they can, this is generally also true for the terrestrial plants.
95% of the ornamental plant production grown in the USA are fertigated(nutrient sprayed), no soil in the roots(bark and sand etc).
Many aquatics have big roots because they do not live submersed all year long or come from high current regions where flooding is common.
Anubias, swords, Crypts, Crinums(some) all fall into this in/out of water group, so large roots help in that case, and from being yanked away, Crypts and swords also store large energy reserves in their roots, same for Aponogetons which are also river plants.
Regards,
Tom Barr