Here is Weiler's transliteration table from Part 3:
a = a or ei or el
c = c
e = e
i = i or l
ii = n or u
ai = ai, en, eil, eli, eii or ell
aii = an, all, au, ail, ali, ein, eni, eui, eiu, elli, elil, eili
The glyphs that occur very frequently at the ends of tokens already do not accord well with common Latin endings or common Latin words (daiin is very frequent in the VMS) but I assume there's more since this only lists a few of the glyphs.
Also, since there are many options for these glyphs, I'm assuming the other glyphs will be handled similarly, which introduces significant degrees of freedom and increased subjective interpretation. Even without seeing the transliteration for the other glyphs, I already suspect Weiler is describing a one-way cipher.
a = a or ei or el
c = c
e = e
i = i or l
ii = n or u
ai = ai, en, eil, eli, eii or ell
aii = an, all, au, ail, ali, ein, eni, eui, eiu, elli, elil, eili
The glyphs that occur very frequently at the ends of tokens already do not accord well with common Latin endings or common Latin words (daiin is very frequent in the VMS) but I assume there's more since this only lists a few of the glyphs.
Also, since there are many options for these glyphs, I'm assuming the other glyphs will be handled similarly, which introduces significant degrees of freedom and increased subjective interpretation. Even without seeing the transliteration for the other glyphs, I already suspect Weiler is describing a one-way cipher.