So for any one else wondering, I went looking for receipts and it seems to check out. Source material claims to be affiliated with the University of Oxford. The database is difficult to navigate, but i was able to find this link to the 4.08.16 english text.
It's honestly astounding how many cuneiform tablets (and fragments of tablets) we have. Multiple ancient libraries full of tablets have been excavated. Now cuneiform was just a system of writing like our Latin alphabet, so they are in all sorts of languages, but we know so much more about those cultures than others of the time because they were writing on clay, then they baked the clay. That makes it last.
They won't. There is some research into long term data storage, though. DNA can be recoverable for almost geological time periods without any special facility providing an optimal environment. There is some work on encoding information directly to DNA.
No. I plugged in a Quantum Fireball drive (~1997, about 9 GB), (IDE to SATA3) bridge and tried to extract the data from it.
The drive platter promptly crashed into the head, the platter shattered, and then a full short began drawing maximum amperage and melting the IDE slot.
So the platter blew up and the drive caught (indirect) fire.
If this interests you, sci-fi author and astrophysicist Gregory Benford did a deep dive into the subject some years back after being put on a committee to try to decide how to mark a nuclear waste storage site as unsafe even if humanity collapses and written language is forgotten.
I think there’s a huge difference between a knowledge repository and a nuclear waste storage facility immediacy is a major factor here as well as the sample size of writing