How perfect is Voynichese?

[ololololo]

For example, one possible explanation of the missing q in the first 1.5 pages of the manuscript could be that some problem (ambiguity or just inconvenience) was identified and fixed by adding a new character.

It would be interesting to identify this problem... In the context of my theory, I could assume that q means some kind of prefix, such as con-, to make it easier to identify and faster guess the right word in the anagram (for example, the word concado. If I align it in alphabetical order, I get ACCDNOO. I can put a part of con "out of brackets", and it turns out CON ACDO. This way you can guess the word CONCADO faster).

Whether they are just decoys is not important here, but the author certainly had some imagination.

I think it is. But considering that even though he was smart, he didn't understand cryptology, it still points to the thesis of this post.

And, conversely, I find it less likely that a 240+ pages long manuscript was (seemingly) completed while using a badly designed system.

The number of pages does not mean that the cipher must be complex or "perfect". The Codex Seraphinianus has about 360 pages, but this does not mean that it is written in a clear and meaningful language.