The arguments I've heard about tracking etc are misguided and don't understand the actual risks.
Firstly, posts on the fediverse are already likely being consumed by advertising platforms like Facebook & Google. It would be trivial for big tech companies to setup relays that act as scrapers.
Secondly, the value in platform's tracking individuals is for advertising. There is no mechanism for these platforms to identify you browsing the we if your instance federated with threads. Your instance won't share cookie sessions etc with threads. It doesn't increase your exposure.
Thirdly, these platforms have the know how to deal with spam and they will be incentivised to share that tech with other federated instances.
Don't get me wrong, Facebook is an evil company. But I haven't heard a decent argument as to why them joining the fediverse is a bad thing. We always have the option to defederate in the future.
Change my mind.
Sorry, but I think you're missing the main point.
The risk is not to be tracked, the issue is embrace, extend, and extinguish (EEE): https://en.m.wikipedia.org/wiki/Embrace,_extend,_and_extinguish
They are currently competing with Twitter and Bluesky, they just need users to kickstart their new platform. That's where the fediverse comes in. All Meta has to do is to convince the instances to give them users.
Meta has a lot of money to throw at UX, they will design a better one than Mastodon. Their instance will also be more reliable (since they have money for lots of computational resources). This will allow them to spread their influence on the fediverse (so that people follow others on Threads), growing up to be the largest instance, and then just defederate from everyone else to “stop spam”. People will then move to Threads so they keep following their friends there (because their friends signed up for meta, since it was all compatible anyway).
And only then, they will start to harvest data and put ads in front of you.