What I’m saying is there are already dashboards (Grafana) that monitor nodes 24/7/365 (Prometheus), and send alerts for any issue (Grafana). They ship with a default set of reports (example), and instructions that many SNOs have used to set them up.
I’d look at those default reports to get ideas about what others think is important to track. But obviously you can add your own spin and polish: can’t wait to see!
(The free Grafana+Prometheus combo is popular: and can be used for almost any app. And they both run well in Docker containers: so if you understand how to run Storj … running that combo is going to feel familiar! )
Edit: I forgot: you can look at the Multinode Dashboard for ideas too! Though it’s more of a rollup of raw stats.