Or maybe it's because the internet is tendentially becoming just a few companies' data centers? Afaik Twitter moved to GCP a few months ago. Maybe this is another Google outage?
Something like this is my bet too, there was a recent post somewhere called something like "why all outages are due to a configuration change". There are monocultures in site reliability ops for big companies, "configuration over code" but with heavy automation too. From my outside view it seems there's a tradeoff when you do that between more frequent smaller issues and less frequent bigger issues. Also reminds me of Google's move away from eventual consistency because with their infrastructure they can make a CP system highly available in practice... except when it isn't, due to a botched configuration change.
It's apparently a word but I'd say it's quite uncommon. I played around with google ngram viewer and had a hard time coming up with a word that is less common. But I finally came up with "astrophotographic".
