Voynichese is a numeric cipher?

[ololololo]

with index tabs sticking out for each initial letter or combination, like "BA-BL", "BM-BR", "X-Z", etc. 

Well, I still don’t really understand how it works (according to this logic, vords are plaintext words, or what?), but in my opinion, such a book cipher could explain the “synonyms”:
   
   

This could have been an attempt to save time (you could have taken an already filled‑out card and added a new word).

Or, if there were several such sets of cards (this could explain the differences in the languages. The sets could differ not only in their lexical composition but also in the meanings of the words. Let’s say that in Currier A, qokedy means “tea,” while in Currier B, it means “blood”), this could be precisely the result of that.

Overall, a single author could use several sets of cards. This doesn’t make deciphering any easier for them, but it does save time—both when creating a cipher (you can take an already used word and assign it a meaning) and when writing a book (you won’t have to come up with a new set of words for a new section).

And overall, it’s expected that this cipher will evolve as the book is written.

But in this scenario, the book could only be used by its author. The intended recipient would face great difficulty in deciphering all the words, which, given the multitude of meanings, would take a great deal of time. This brings VMS closer to the Rohonc Codex, which we now know could only be read by its author (well, maybe also his supposed deaf teacher, but that’s debatable ).