Why not positional variation?

[dashstofsk]

This would require showing that these words are language-dependent.

That is assuming there is language dependence. It might be that the manuscript is in no language, a meaningless fabrication. My approach has been to do many statistical analyses on the text, then look closely at the results to decide which of these hypotheses is the more probable. Unfortunately, my analysis of iin, in does not seem to me to be consistent with the meaningful language hypothesis.