This question is a follow-up to this question Latin passive endings: Why is -mini sticking out. The Latin 2nd person plural passive ending mini has attracted the attention of scholars for centuries, but no satisfying explanation is found so far.
Since the r-mediopassive goes back to Proto-Indogermanic with attestation in other branches than Italic and Latin proper, does the outer perspective shed some light on this ending? (EDIT:) In particular, was the apparent irregularity already present in the proto-language, or is the Latin ending -mini a later replacement of something more regular?
I have looked into Old Irish, but alas, it has already simplified the passive paradigm featuring only one ending for all plural persons. Therefore it is of no help here. Do other Celtic languages or Tocharian give some new insights?