Why and how the text could be Bavarian

[JoJo_Jost]

Your method is ingenious and surely interesting (and I think there have been few studies which considered spaces not to be true spaces). I always assumed spaces to be real spaces, if not because they are the most frequent character in the VMS, just as in every other language. But equating spaces to vowels works too, because after eliminating the spaces from a text the remaining most abundant characters are vowels.

Thanks

A problem I see is that, if I understood correctly, the cipher will always encrypt the same plaintext string to the same encrypted string, which should create many cases where the same Voynichese words repeat in the text, which is not what is observed.

That's a very interesting question, and it's essentially correct. 

If you only transcribe the vowels, then it would work:

Let’s define
e/ea = y[space]qo
i/ie = y[space]ch
o/oi = y[space]sh
a = y[space]o

Different variations of “Home”
y othy shmy qo... = at Home
y chnhy shmy qo... = in Home
ty shy shmy qo... = to Home
fry shmhy shmy qo... = from Home

(btw: here u see, why Line Start Markers are needed, like in VMS)

The consonant-vowel cluster “hy shmy qo” would therefore theoretically be the same and would have to repeat. I think that’s what you mean.

However, everything after “qo” would depend on what comes next;

But it’s not just the vowels that are encoded—the encoding is somewhat more complex:

As already mentioned, internal consonant clusters are encoded differently.

The “hy” would therefore depend on which word precedes it.

othy
chnhy
shy
shmhy

These consonant clusters are replaced partly by the E-family and partly by other families (aiin) or others. This, in turn, can also influence the vowel part on the right.

So “shmy qo” or only "my qo" would remain a possible repetition.

What I haven’t written yet, because I haven’t quite deciphered it myself (i have some ideas):

It must be the case that when a word ends on a vowel (like "home" here) and a new one begins with a vowel, a specific cipher is applied again—precisely because two vowels cannot be encrypted consecutively. In that case, the “y qo” would also be omitted in these instances in this example.

What would remain is the "shm" or “m” as the core of “Home,” which is repeated—and that would no longer be recognizable as a repetition.

In VMS, bigrams and trigrams do, in fact, recur quite frequently.

But i'm working from the Bavarian dialect, and it is an almost monosyllabic language, there aren't that many longer words. Many Words are: = Consonants Vokal Consonants = "CVC". If "CVC" meets "CVC" then you have CVCCVCCVCCVCC And the CC Cluster are part of the E-Family and others. So there is no repeat, too.