Deconstruction of Voynich Manuscript words

[Emma May Smith]

Many of the strings of glyphs in different groups seem similar but for one small difference. So you strings containing [k] and [t] which are the same except for that letter. Given that [k] and [t] are so similar and can be thought of as a class of character (gallows), why not simplify your process even more by merging those groups of strings and replacing [k] and [t] with a generic G for a gallows character?

Likewise, which list strings beginning [o] and [q] separately? Surely it would be much easier to say that any string beginning [o] can optionally take a [q], which is roughly true?