A decent setup here should survive 3 sessions without looking messy. The journal should show whether the numbers are real.

tbh the value is mostly in filtering bad ideas. I still would not trust it live without checking the chart myself. The useful comparison is hit rate versus average R.

For trend filtering, this feels more useful as a filter than a full system. The journal should cover 69 setups and the trade context.