Why not decouple the community and metadata from the content to protect from losses like this?
The wealth of the metadata present in what.cd that represents years of work far surpassed anything available in public 3rd party sites.
And there you get an IRL banhammer.
A lot of torrent "trackers" already only store the content ID on the DHT, it doesn't stop them from being shut down...
You can find more info on how to install and run it here: https://privacytutorials.wordpress.com/
rutracker.org is hosted in Russia.
libgen and sci-hub are using a more complicated setup with lots of fronts (in various countries) run by people who know what they're doing. I believe the backend infrastructure is in Russia.