Finally, a clear explanation of Zettelkasten.
Textmind differs in that it doesn't rely on tags and manual linking. Instead it uses deep hierarchy.
How did Luhmann make his Zettelkasten work as a Personal Knowledge Base (PKB)? He didn't. He was writing on academic topics for publication. Tags can work in that context, because they are essentially for collaboration. Wikipedia is flat but works; it doesn't need hierarchy. The namespace isn't overly cluttered because the scope is general knowledge. The manual interlinking isn't overly onerous since it is intended for publication and therefore write-once, read-many.
But for a PKB, an evolving hierarchy is necessary, to reflect the user's private nomenclature with which he runs his life. The read to write ratio may be as low as 1:1. It must be fast and inductive to allow the user to name all the issues in his life. Zettelkasten lacks inductive nomenclature discovery and thus can't help.
A PKB must manage projects, contacts, objects, etc. It defies the assumptions which permit collaborative schema to function. Deep hierarchy is the only way to disambiguate the many shades of meaning each "tag" can have. And rapid refactoring must keep those meanings synced with evolving circumstances.
A Zettelkasten is great as a pre-publication tool. It enables discovery of interconnections between publication-ready atoms. Today there is no reason not to publish the Zettelkasten itself, and some people do so. Hence the principle: Zettelkastens are inherently collaborative.