Is there any character like that?

[ReneZ]

For me the more important sign is not how small the variation is, but how unambiguously it splits letterforms into distinct groups.

I like this formulation.
Now I know that @nablator has spent a very considerable time transliterating the text and looking at exactly these variations.  His comment about a "continuum" is spot on, in my opinion.

A short summary:
It is not possible to unambiguously split between (non-exhaustive list):
o and a
r and s
Ch and Ih
m and g

In fact, in all cases, it would be possible to define a form in between each pair, but that only makes matters worse: instead of one fuzzy boundary we end up with two (for each pair).

Back to the point: these pairs are already considered to be different characters, but they are part of some continuum. This is even worse than the situation that each of these individual characters has a continuum of variations.