"Once in a blue moon" -- you mean the thing that constantly happens? If you're not using logs, you're not practicing engineering. Metrics can't really diagnose problems.
It's also a lot easier to inspect a log stream that maps to an alert with a trace id than it is to assemble a pile of metrics for each user action.
It's also a lot easier to inspect a log stream that maps to an alert with a trace id than it is to assemble a pile of metrics for each user action.