Yes, I get that but wouldn’t launching new features be adding pressure during an outage which complicates the things? Capacity issue also could have been caused by a new feature probably acting greedy on the resources too.
While true in principle... those kinds of causes tend to have pretty clear cause-effect (or at least strong time correlations), and get rolled back. New call patterns / major load changes are generally pretty obvious and at least coarsely traceable if you have monitoring anywhere.
