The Cardan grille approach to the Voynich MS taken to the next level

[Anton]

The topic of the paper is individual word generation, so the (problematic) issue of word repetitions, and lack of longer repeating sequences is outside.

If so, then it would be more correct to speak of generation of the VMS-like vocabulary, not the VMS-like text. Because text is naturally a certain sequence of words, not just an unoredered set thereof.

With respect to label morphology, I am not aware of any 'issue' (i.e. problem).

I mean e.g. the abundance of "o plus gallows" as label-start.