Because when I was young we were having phones, even mobile phones, but first few years - without internet. My mobile site was one of the first mobile sites in Russia. And I know for sure: communication by phones/mail is as different from internet communication, as planet is different with atom. Coordinated networks are MUCH different than list of contacts in phone.
Stopped reading after story about cron script killing long running db queries - it means this system is a huge mess, where killing request is more simple solution than finding which part of code generates it. I saw such systems and I'm sure it's hopeless state and only as sarcasm I can read "thankful to architecture principles" here.
Why not both? It's not unheard of for a deploy with index issues to cause a load spoke in production. Yes, you need to know about the query to fix it, but also killing runaway queries agree they're obviously orphaned can mean the difference between an outage and a slight site degradation in response times.