Matrix is made large enough that the inside of the rocks will be anaerobic even without an oxygen gradient. like a small peice of live rock after a while. That being said, I do think what you're discribing could work better. With depleated oxygen in the tube (increased by a slow flow rate) could cause more of the matrix to be anerobic compaired to in fully oxygenated water. Problem with this is the inlet would have to have filter floss to prevent most of the clogging of the matrix. It woun't get clogged and stop flow, but it will collect detritus. I keep a layer in a canister filter like a reactor and it seems to work well, however every other week I clean it out and rince the matric in used tank water to remove any build up.
If you're simply looking for nitrate reduction, i'd also look into a remote deep sand bed if you havn't already, or carbon dosing if you don't already do that. I have a deep sand bed in my tank and carbon dose and can very easily strip my water of all nitrate if I wanted. i purposly feed heavily to keep nitrates up.
