I did not plan to announce it yet as it is in a very early stage, I was targeting November, but I think it should be mentioned now: I am working on OpenWeb, https://framagit.org/open-web/open-web, a project which allows to access posts (at the moment only tweets but more platforms to be supported later) without any account nor tracking. It *downloads* them instead of proxying, with the goal to make the data public data.
Of course, it can't download the full twitter database, so the scope is limited to entities which *have a wikipedia page*. That's the criteria for "is it famous enough", as wikimedia already has good rules about who/what deserves a wikipedia page or not.
More on that project soon.