Customer problem: simple averages hide setup time and produce weak quoting data.

We're considering a better way to separate setup time from per-piece cycle time so quote feedback and production learning are more trustworthy. This is especially relevant for shops where setup-heavy work distorts simple averages.