So, I'd just made that suggestion myself a few moments ago.
That said, there are concerns with data aggregation, as patterns and trends become visible which aren't clear in small-sample or live-stream (that is, available in near-time to its creation) data. And the creators of corpora such as Twitter, Facebook YouTube, TikTok, etc., might well have reason to be concerned.
This isn't idle or uninformed. I've done data analysis in the past on what were for the time considered to be large datasets. I've been analyzing HN front-page activity for the past month or so, which is interesting. I've found it somewhat concerning when looking at individual user data, though, here being the submitter of front-page items. It's possible to look at patterns over time (who does and does not make submissions on specific days of the week?) or across sites (what accounts heavily contribute to specific website submissions?). In the latter case, I'd been told by someone (in the context of discussing my project) of an alt identity they have on HN, and could see that the alternate was also strongly represented among submitters of a specific site.
Yes, the information is public. Yes, anyone with a couple of days to burn downloading the front-page archive could do similar analysis. And yes, there's far more intrusive data analytics being done as we speak at vastly greater scale, precision, and insights. That doesn't make me any more comfortable taking a deep dive into that space.
It's one thing to be in public amongst throngs or a crowd, with incidental encounters leaving little trace. It's another to be followed, tracked, and recorded in minute detail, and more, for that to occur for large populations. Not a hypothetical, mind, but present-day reality.
The fact that incidental conversations and sharings of experiences are now centralised, recorded, analyzed, identified, and shared amongst myriad groups with a wide range of interests is a growing concern. The notion of "publishing" used to involve a very deliberate process of crafting and memoising a message, then distributing it through specific channels. Today, we publish our lives through incidental data smog, utterly without our awareness or involvement for the most part. And often in jurisdictions and societies with few or no protections, or regard for human and civil rights, let alone a strong personal privacy tradition.
As I've said many times in many variants of this discussion, scale matters, and present scale is utterly unprecedented.