chore(planning): calibrate S M L XL effort estimates against historical PR sizes

[ohdearquant]

Body:

• Problem: roadmap effort estimates are currently S/M/L/XL and were not calibrated against historical PR sizes, which weakens schedule confidence.
• Proposed fix: compute historical PR size and cycle-time buckets for comparable khive changes, then publish a local calibration table used by the roadmap.
• Acceptance criteria: calibration table maps S/M/L/XL to observed changed-file and line-count ranges; at least three historical examples support each non-empty bucket; roadmap estimates cite the calibration artifact; outliers are documented.
• Affected crates/ADRs: /Users/lion/khive-work/shows/khive-improvement-survey/feature-frontier/synthesis.md:94, analysis.md:97.

---

Source: khive-improvement-survey show — proposed backlog (propose-only synthesis). Dedup: NEW.