Again im no conspiracy theorists but I feel like this yes can happen by accident but many many things have to go wrong and for it to be out for up to 24 hours means even more things had to go wrong. It just feels wrong is all..
So for other non geek people BGP stands for Border Gateway Protocol and it's like a lookup table that when you say go to pokerfoum.ca that site name is converted to an IP address and data is sent to that IP address out through your modem to Rogers network. It will reach a Rogers switch the and that switch will use BGP to decide what network it should go down to get the data to where it needs to go to efficiently. Rogers could have a Netflix caching server in one of its locations so all rogers switches would have a BGP route in their tables to route all Netflix traffic to their internal Netflix caching server so you dont have to send your data all the way to New York or California. This makes things like Netflix faster when using Rogers. There are also fallback lower priority routes incase say that caching server is down or that network segment is down. If there is any delay or reason a sending switch does not get an acknowledgement of a packet received from a route then that route gets pushed down in priority and another route is tried and eventually a route can be declared not useable or down etc.. I am very much simplifying it here but it is a simple protocol and its why its used and very powerful.
Here is a great article on it from cloud flare
https://www.cloudflare.com/en-ca/learning/security/glossary/what-is-bgp/
Not sure why it takes 24 hours to put back simple records like these but i also dont do IT networking on a scale like this! no idea what it must be like to jump start a network like this from dead!