In any case, I'd like to stress the point that is trivial but nonetheless easily forgotten in a discussion about the possible rules or algorithms.
If one suggests a procedure for encoding, then a procedure for (unambigous) decoding should be conceivable as well.
And, vice versa, if one suggests a decoding procedure, then the respcetive encoding procedure should be conceivable.
Respectively, if a set of "instructions" is supposed to convert Voynichese to plain text, then what would be the set of instructions to achieve the opposite? How, with a specific output, we achieve "[qo present]+[any galllows]+[ch]+[any group from column 6]+[o]+[any group from column 11]"?
If one suggests a procedure for encoding, then a procedure for (unambigous) decoding should be conceivable as well.
And, vice versa, if one suggests a decoding procedure, then the respcetive encoding procedure should be conceivable.
Respectively, if a set of "instructions" is supposed to convert Voynichese to plain text, then what would be the set of instructions to achieve the opposite? How, with a specific output, we achieve "[qo present]+[any galllows]+[ch]+[any group from column 6]+[o]+[any group from column 11]"?