The logic you'd love to see communicated automatically is too complex. In addition, there are always other reasons (change of mind, drunk operators, ...) that don't fit into this. I, personally, prefer the rating system implemented in lnd-manageJ (see https://github.com/C-Otto/lnd-manageJ/blob/main/rating.md), but that's not the only information I consider.
I'd consider your script a well published set of guidelines that I could reverse into a script to run against my own node to see what score you would give me. So maybe I'm not looking for a lightning standard so much as a best-practice for operators to publish their general standards/scoring methods. IMHO, the best metrics are ones that can't be gamed without the metric being realized in a positive way so publishing these things is good because it helps to validate that you really do have a good rule set rather than an arbitrary hack of drunk decisions ;)
reply