Why and how the text could be Bavarian

[Stefan Wirtz_2]

Here’s how it works: The double consonant (usually) appears after a short vowel. A good example: "Ratte" (rat)-short "a," but the "t" is pronounced like a single "t." If you drop one "t" you get "rate" (English: "guess"), which is a verb. In that case, it’s a long "a."

You don‘t need some verb here:
german „Rate“ (= engl. „rate“, of interest, growth or anything) and german „Ratte“ (= engl. „rat“, surely not interesting rodent) are 2 completely different nouns, just recognizable by the length of „a“: 
long in „Rate“, short in „Ratte“. As a result, the longer word is spoken a bit faster — which is quite the opposite than having a short break before doubled consonants. Not the „tt“ causes a length, the „t“ does it by keeping the „a“ long.

@Stolfi: your example „Krabbe“ is spoken nearly like [c]rubbe[r]; short „a“ like short „u“ in engl. rubber.
There is no German word like „Krabe“, but a close one, „Knabe“ (old-fashioned word for „Boy“). Not surprisingly, with a long-pronounced „a“. This word is also a bit longer spoken than „Krabbe“:
you all may decide by yourself, whether the „bb“ sounds like a „b“ and the „tt“ sounds like a „t“ or not.

Problem is, that German authors used randomly doubled consonants for words we don‘t expect there.
This is an example:
   

„Lufft“ (engl. „air“) and „Meisterschafft“ (engl. something like „mastership“, not really „championship“)
are written with „ff“ for not any reason. „Lufft“ can be found from medieval scripts on to books of 19th century, it was standardized to today‘s „Luft“ then, as it cannot be messed up with a similar word.

I don‘t know of any book or script that ever came completely without any doubled consonants. VMS would be the first and only one: it may be really unique, but not in this way…

And that, of course, makes a difference.
But, to be clear: No cipher in the world would encode German double consonants; that would be the key to direct decryption. In a phonetic cipher, there wouldn’t even be a double "t" 

To be clear: without clear indicators there is not any chance to separate „Rate“ from „Ratte“.
Some „just phonetic“ script can never exist due to this structural problem.
@JoJo_Jost: you just confirmed here that „phonetic bavarian“ cannot work.

@Aga Tentakulus: not quite, „facial expression“ would be „Miene“.
„Mine“ keeps to be engl. „mine“, delivering coal or gold, or a heavy explosion.
And, too sad, Alemannic is just a small fraction of German, it‘s rules are not „overall“ rules.

edit: I see there was a comment about stenography meanwhile:
well, I had not to learn this anymore, but I know it as a complex system of abbreviations for syllables, words, and parts or complete (often used) sentences. Stenography exceeds the few symbols of Voynichese by numbers -
if you try to put this as something like a „phonetic stenographic bavarian notation“: good luck, you have at least one foot in the tar pit then already.