I am exited to see, this moving along. We do have several competing systems right now:

1. there is MediaWiki:Gadget-ImageAnnotator.js a javascript code developed in 2010 by user:lupo and rewriten in 2019 by ESanders
2. we also have Help:Extension:FileAnnotations a php code by User:MarkTraceur from ~2016(?), which is a "improvement on the Gadget-ImageAnnotator"

It is unclear which is used by Template:ImageNote with 196k transclussions, but I assume the first one.

We also have SDC based approach utilizing relative position within image (P2677) property as qualifier to depict properties. There is Wikidata Image Positions tool written in 2018 by Lucas Werkmeister. See description. It would be nice to integrate the two somehow. Some of the issues I see:

• license issue: Wikitext on commons is CC-by-sa and SDC is CC-zero, so mass migration of current annotations might not be possible
• the current annotations consist of wikitext. They might include: link to another image, link to wikipedia article, wikidata item, text of inscription, comment like "nice hat", or any else. SDC annotations are much more structured, we can add P2677 qualifiers to depict statements or inscription (P1684) statements. But we might not have a way of storing essays about different details of an image.

It would be great if we could move some of the existing annotations to SDC (like in this file) and alter the code we use to add anotation to pull anotations either from wikitext or from SDC.