There's plenty to criticize about DCNH, and I view it as a inferior subtype system in practice to the 2 subtype systems, however I think there are observations in DCNH that are not noticed in the various 2 subtype systems. I try to deal with the specific problems of DCNH rather then ignore it or try to make it a popularity contest.

DCNH deals with mainly roles people adopt in interaction due to function accentuation and strengthening and it is based on observation of identical interactions, it allows further more detailed compatibility models by hypothesizing how different versions of identicals and other types(and their subtypes) can mesh better and more cohesively.

That has significant usage imo and is a significant observation that doesn't exist with the 2 subtype systems(which are also incomplete in their own ways). Typing is unreliable, sub-typing even more so, nothing new here. There are people that can deal with this and others that pretend to deal with it by ignorance and/or overconfidence. You see this happening a lot on the forum and in general in many topics. It's clearly a hypothesis at the moment, but imo there are definitely observations being made which are legitimate and found no where else in socionics at this time.