This article suggests an interesting theory, but the author didn't actually test it. He was only sharing something he found elsewhere and even that source seems to only "suggest" what might happen. Personally, I doubt that much, if any, anaerobic bacteria develops (to reduce nitrates) in an UGF because the water movement is unlikely to be slow enough and, I would think, any water moving completely through the substrate is going to be oxygenated. Additionally, you are supposed to deep-vacuum regularly, to keep flow going.
Anecdotally, my experience was that nitrates climbed as a function of the sludge buildup under the UGF. I was able to remove about half that sludge by inserting a flexible tube down the flow tubes and suck out some of it (a difficult process). After doing this, nitrates would drop, but we are only talking about 5-10ppm. As I mentioned, above, I found that by adding a HOB, I was able to reduce sludge buildup, somewhat, and keep nitrates lower but, perhaps, the HOB was simply adding more stabilization to the overall filtering need. I finally bit the bullet and pulled the UGF out. This further reduced nitrates. However, I will admit to habitually overfeeding my overstocked tanks and, as you know, this can contribute to sludge problems.