We would like to track medium-sized solutions perf as most of the smaller improvements/regressions would not be visible on smaller scale.