When I decided to dive into VMS, the first thing I did was to read/scan the pages over a few days to understand its characters composition. See attached image.
From it, I believe that
ee and
ii could each be a single character. (And thus,
eee [and
iii] has to be single character too or else there would be a situation like, is it '
e ee' or '
ee e' or '
e e e'?)
The tails of
b, n, r, s, l, y are just curved tails, regardless their 'shapes'; they could be tall, short, wide, long - and the pages are abound with their variations which I believe were dependent on the style and mood of the scribes who 'drew' them. The connection point between the tail and the 'base' character
e or
i could be made at the top or bottom of the tail, with the top or bottom of the
e i character, except there is no connection between top of the tail to the bottom of
e i character as such connection would made the new character's descender far too low. I believe the macron in
c"h is also a tail, but scaled down to fit into the limited space available.
The gallow characters are variations of the right two columns (
g m), but use a tall vertical stroke instead of
e or
i. The character
q is probably derived from
e and a vertical stroke downward, and
o is a special character (possibly with special attributes).
Just sharing my 2 cents worth....