Me
ERA is so distorted for relievers. 9 one inning stints without giving up a run or hit followed by an outing with three straight walks and get pulled. Next batter hits grand slam. You're charged with 3 earned runs and 0 innings pitched. Your season ERA is now 3.00 despite not giving up a hit. Just not a great metric for relief pitchers.
Giving up 5 runs in one outing shouldn't define effectiveness of a reliever.
Notin
As Adam Ottavino showed last season...
Lets think outside of box.
ERA is more appropriate for starters. If you go back into history, it was common for starters go 9 innings. Projecting starter innings to 9 innings make sense to me. Basically it gives you an idea of how many earned runs are given up by a starting pitcher. Maybe this was the genesis of ERA. (earned runs per nine innings)
Not so fast for relievers in modern age. Many are one inning specialists. No one gives a damn about 9 inning projection. Relievers don't pitch nine innings. (exception of Eovaldi).
For relievers, I would simply take total earn runs given up divided by number of appearances with one modification. Earn runs given up in any outing exceeding one inning would be adjusted, pitcher's earn run would be runs given up divided by innings pitch. That's the run given up number for that outing. (ie restate the runs given up per one inning basis). So you come in and give up up 3 runs in 3 innings, your number would be 1.0 for that appearance.
Reliever's stats would be called Earned Runs per Appearance.
In above example, his 'ERA' would be 0.3. He gives up .3 runs per outing.