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

[ReneZ]

If the wheels operate independently it would be easy to split the words into independent parts.

It wouldn't be that easy, among others because of the great uncertainty related to word spaces.
Rare combinations of the parts can correspond to rare words in the plain text.
Typically, for a text of this length, about 15% of all word tokens, and over 50% of all word types are Hapax, i.e. words that appear only once in the text. (Rough figures).

Stolfi is splitting common Voynich words like daiin or qokeedy into two parts. Tiltman also splits Voynichese into two parts. Didn't that mean that we should assume the existence of only two wheels?

As described in the paper, Tiltman's table generates only 240 different word types, out of ca. 9500 needed.