The truly worrying thing about the Rogers meltdown is if there is anyone competent at Rogers and they were listened to it would not be possible for all of the infrastructure of a telco, encompassing Internet, voice, private data and so on to collapse together.
The only way this can happen is by creating a single point of failure for a collection of systems that, if built organically, or haphazardly, or even randomly, would not naturally have a single point of failure. You have to really go out of your way, and be monumentally stupid, to create a scenario where this is possible.
My whole life in the ISP/telco world was about building in as much fail-safe, redundancy, and generally avoiding "house of cards" scenarios as ingenuity and budgets would stretch to. i.e. the exact opposite to what Rogers must have done.
I've presided over some nasty outages in my time, but the longest one I can recall that affected multiple customers simultaneously was about 4 hours, and was restricted to just one area, perhaps one Internet PoP, or a set of International voice routes, etc. etc. I never had a whole network fuck up that took out all services and the worst, as in largest number of customers affected, would have BGP related, taken 15-30 minutes to fix and perhaps two hours for everything to settle back to normal.
Incredible, truly incredible.
There's been a trend recently for larger and larger service providers to somehow manage to cause whole network crashes which affect millions of direct and indirect customers. That's worrying in itself, but no one seems to have managed to fuck up as comprehensively as Rogers appear to have.