Probably never going to find out as this is not the type of stuff they're release to the public, but I'd be so curious to know what the issue was, from a technical standpoint. It had to be something very major related to something centralized. I'm thinking my theory of route table issue can't be it at this point given how long this lasted.
Must be some central piece of equipment that failed, and it was probably hard to get parts for it or something. I'm thinking some kind of transport equipment or something though typically that would only affect a certain geographical area not the entire network.
It is crazy though how fragile the internet and overall communication network really is. Often times we think of it as a big mesh, but it's not actually quite how it works. There are actually lot of single points of failure throughout the system.