At first glance, your argument about the ‘167 prefixes’ seems convincing, but I think it is based on a category mistake.
According to my hypothesis regarding the Bavarian cipher, what appears as ‘prefixes’ before ‘daiin’ are not linguistic prefixes in the conventional sense – rather, they are combinatorial results of a structured absorption layer, which, for example, might function according to a small set of rules:
o- absorbs the article (der/die/das)
qo- absorbs combinations of preposition and article
y- absorbs verb prefixes (ge-, er-, be-, etc.)
Gallows signs encode consonant classes
Vowel clusters encode vowel quality
These elements combine multiplicatively. A handful of structural rules generate many unique surface signs – exactly what one would expect from an absorption cipher applied to inflected Middle High German.
As for ‘daiin’ specifically: according to my theoretical correspondences (d=s/z/ts/ß, aiin=ein), this yields ‘sein’, ‘zein’, ‘tsein’, etc. ‘Sein’ is the most common verb and possessive pronoun in German. But beyond ‘sein’ as a standalone word, Middle High German features an extraordinarily productive -ein-syllable, which occurs in hundreds of compound and derived forms. Here are examples from my current reference corpora (Ortloff von Baierland, Breslauer Arzneibuch, Admonter Bartholomäus):
Cleine
Kein
Keine
Marein
Obeine
Osein
Seind
Seint
Stein
Steines
Sweinen
Wein
Winstein
Weinstein
aitstein
algemeine
alleine
aufeinander
bein
beine
beinen
beinlin
beinwell
belein
blutstein
cbreinerhande
cherndlein
chlein
Zicklein
darein
dein
deine
deinen
deiner
deines
deins
drein
elfenbein
entreinit
erein
erlein
federlein
fein
feins
gagein
gaissein
gaisseine
gaisseins
galizzenstein
gelein
gemeine
gemeinde
gereinigen
gereinigt
geysseiner
glein
grysssein
guldein
hamatstein
harnstein
herein
hiersein
hinein
hirse
kein
keine
keinen
keiner
keines
keinz
keins
kelbrein
klein
kleine
kleinem
kleinen
kleiner
kleines
kornlein
kummerlein
kuppffrein
latein
lateinisch
lein
leine
leineins
leinen
leinsamen
liechtein
meine
meiner
mitein
miteinander
neglein
naglein
pellelein
peterlein
pilein
poneinem
plein
puchkein
quintein
quitein
rein
reina
reindlein
reine
reinegen
reinem
reinen
reinigen
reiniz
ribesteine
ritlein
salbein
schrein
schweine
schweines
sein
seind
seine
seinem
seinen
seiner
semelein
sirein
stein
steine
steinpheffer
steinwurtz
sweine
sweinem
sweinen
sweins
tanneins
tein
teincken
tennein
tlein
unrein
unreine
unreinen
vein
waiczein
wein
weinbeffen
weinber
weines
weinessich
weinic
weinich
weinpern
weinreb
weins
weintauten
winstein
zerklein
zerkleine
zerkleinen
zinemein
160 words from actual medieval texts, and that is likely far from all that is possible.
Each of these forms, when processed by the cipher’s prefix rules, would result in a different character preceding the component -ein. The 167 variants are not 167 independent prefixes attached to a word – they are the predictable result of a small set of structural coding rules applied to the most common syllable group in the target language.
If VMS were a directly written natural language, the argument would hold true. But that has long since been disproved.