I think suggested edits are a great useful approach (also for categories and adding media from WMC). However, currently this seems like a huge time-sink that is not efficient in any way. Does the tool suggest machine translated captions translated from the file description or other Wikipedia articles that the same image is used (especially EN WP)? If not, please add that so that people only to correct or adjust the translated captions which often would be better if they write something new – see tools like {{w|DeepL}} and Google Translate usually also works good enough for texts as concise and short as captions usually are.
Moreover, I think it needs ways to add media to many language versions at once such as including all media set on the Wikidata item, see my proposal here. Suggested edits with AI-support would result in far more contributions in far less time and with substantially higher quality. Also consider that there's over 300 languages, writing the captions for just one image takes a lot of time when considering there's so many languages. The MinT project is also relevant. Lastly, nowhere is it specified which Android app is referred to and what the latest status of this is. I guess it's the Wikipedia Android app right? Is this feature live there, please add some info what this project concluded in, thanks.