We've become so used to good service that we often fail to appreciate how much has to go right for things like the internet to work. The internet is designed to cope with having to reroute traffic but at the cost of latency.
And as the internet becomes more important to us, it also becomes a bigger risk. What this article doesn't say is that carrier failure is a fairly common (not not everyday) occurrence, however brief, in many parts of the world. Prioritising reliable Tier 1 providers is actually one way to mitigate this but even they can have their problems. I've yet to see a CDN that doesn't have the occasional wobble in one part of the world. As with all failures, when it happens good communication is key. Along with working out exactly what went wrong and whether you can prevent that happening again. The question is then: is Cloudflare's and Telia's analysis plausible and do their proposed changes sound reasonable? Or is this more like TalkTalk's crisis management?