I am building a n8n health tracker for my company with these goals:
- Track all executions (success, error, partial success etc)
- Time per execution (wf and sub-wf levels)
- resource utilization
The final goal is to track and make n8n be a solid pipeline with modules than just prototypes. Would love to hear your experience (if any) or just brainstorm around this 
Yes I used the built in data tables to store stats about my workflows as it progressed through different parts and their performance and then created a webhook triggered workflow to read the stats data table and return all rows which I then called from a web dashboard.
You just need to think outside the box a little, you can record times etc easily.
which is okay, but that’s not sustainable for a massive repository of WF and processes (100+ WF+sWF across 15+ processes)