The information is usually not identifying beyond general geographic regions (at best)
if your threat model is that strict, there are other ways you should be obfuscating your IP than relying on VPNs, ISPs, and the apps/servers you're accessing/using.
Usually it is the town, or nearby town. If you live in a more rural area that can narrow it down to a few hundred people.
Also in some less-developed countries the data protection by ISPs is very weak. Basically if you know someone in the police (or pay a bit under the table) you can easy get the exact name and address of the account owner if you have an IP.
Here's a professional security researcher/pentester explaining in depth why "leaking" IP is blown out of proportion
The relevant gist is
Usually it is the town, or nearby town. If you live in a more rural area that can narrow it down to a few hundred people.
Also in some less-developed countries the data protection by ISPs is very weak. Basically if you know someone in the police (or pay a bit under the table) you can easy get the exact name and address of the account owner if you have an IP.
Here is an alternative Piped link(s):
Here's a professional security researcher/pentester explaining in depth why "leaking" IP is blown out of proportion
Piped is a privacy-respecting open-source alternative frontend to YouTube.
I'm open-source; check me out at GitHub.