Plenums were developed and introduced by the Monaco aquarium quite some time back. Such consists of a "dead space" at the bottom of the tank, generally created by a sheet of eggcrate plastic covered with fiberglass mesh screening, supporting some depth of substrate of a fairly fine particle size.
Everyone is or should be familiar with the possibility of anoxic areas deep in the substrate, and if the subtrate has become enriched with organics over time, the posiblilty of dangerous gas production (H2S, methane, etc.) under profound anoxia. But fewer seem to be aware that under conditions of less deep anoxia* that nitrate is reduced to nitrite and on to ammonium and that on to ntirogen gas, mediated by bacteria just as the oxidation if ammonia of ammonia is done in high ORP situations (our filters). The bacteria (anaerobic to anoxic) are quite diffent, and the process is slower to establish (diffusion, not flow, is involved - a much slower process). But it does and can happen and can be used in SW (first use and where it was developed as a practical process), BW, or FW.
BTW, the quote cited was astonishing in its ignorance. There is nothing inherent or magic in a glass or acrylic box that forces full-natural processes to stop at some arbitrary stage. Ammonium certainly can and is routinely reduced to N2 gas in many hobby tanks (requiring only deep substrate) whether the tank operator is aware of it or not. It is absolutely astounding to me that some people think there some sort of magic involved in tank process involving biochemistry. We are a community of the biologically and biochemically naive, and as such seem prey to mind-boggling mythology. If the conditions are right for nitrate reduction, they are right for nitrite reduction, and for ammonium reduction. Phosphate handing is a separate issue, as is H2S or methane generation - separate condition requirements exist for each, which is why we can take and use the mildest (highest ORP) case (nitrate to N2) and use it if we wish. Additionally, if an energy source exists at some depth in the substrate, bacterial species which can utilize that source will colonize that area/depth/environment. We see that in our high-ORP biofilters. It may be slower than we would like, but it can and will happen. The movie tag "If you build it, he will come" can easily be applied to the microbial world - if a food source exists or develops, users of that source will come and will establish. Every region of NNR is capped by regions of users (oxidizers) on ammonium and nitrite. Certainly the intermediate products, nitrite or ammonium, diffuse from the area, but if they diffuse upward, they are consumed by the higher-ORP bacteria at that level, and thus recycle, just as they do in our aerobic biofilters. No magic is required or involved.
*Anoxia and or anaerobic conditions are not a single state, but a continuum. All substrates have a decrease in oxygenation with depth, controlled by the particle sizes in the substrate and by depth. Very fine particle substrates have a very steep decrease in ORP (oxidation-reduction potential) with increase in depth. Larger particle substrates have less steep declines. Below a tiny depth in either, diffusion becomes far more important than flow, and smaller particles hinder diffusion quite well - there are fewer and smaller more torturous channels than with larger particles. This is responsible for the risk of undisturbed small-particle substrates.
With sufficient depth, any particulate substrate will have levels of reduction in oxygenation with increases in depth. At some depth the environment will be appropriate for normal denitrification, the reduction of nitrite to nitrite to ammoniun to nitrogen gas (N2). Still deeper, the O2 tension/ORP will be still lower, and allow for sulfide and iron reduction, and even methane generation.
The brain at Monaco got the concept of controlling the depth of anoxia possible deep in a substrate by placing a plenum (literally a vacant space) under the substrate. Vacant spaces - water only - do not hinder diffusion at all. this would in theory block the deep anoxia (very low ORP) required for the nasties, but allow what came this side of the pond to be called NNR (normal nitrate reduction) but blocking methane, etc. production. It works.
Note: Undisturbed deep anoxia or moderate anoxia are neither harmful in and of them selves. Methane and H2S generated in very deep and fine substrates will be oxidized as they diffuse upward, as the overlying levels will be colonized by bacteria which can and will oxidize them. Ditto for nitite and ammonia from NNR areas. The danger is in and if they are disturbed by inhabitants or by the hobbyist. Then toxic material can be released, and the critter injury or death can follow.
HTH.