feat: add /api/stats analytics endpoint

[AdityaAsopa]

Summary

Adds a new /api/stats endpoint that returns aggregate statistics computed at runtime across all five data collections.

Response shape

{
  "totals": {
    "spacecrafts": 113,
    "launchers": 81,
    "customer_satellites": 75,
    "centres": 44,
    "spacecraft_missions": 64
  },
  "spacecraft_missions": {
    "by_status": { "active": 3, "decommissioned": 44, "failed": 3, "unknown": 14 },
    "by_orbit_type": { "GEO": 22, "SSO": 12, ... },
    "by_mission_type": { "Communication": 15, ... }
  },
  "spacecrafts": {
    "by_status": { ... },
    "by_orbit_type": { ... }
  },
  "customer_satellites": {
    "by_country": { "USA": 10, ... },
    "total_mass_kg": 84235.2
  },
  "launchers": {
    "by_vehicle_family": { "PSLV": 30, ... }
  }
}

Design decisions

• Nothing hardcoded — every number is derived by iterating the JSON arrays at request time. The response stays accurate automatically as data files are updated.
• Null fields counted as "unknown" — consistent with the normalization approach in Normalize data schemas, enrich datasets, and fix API handler quality #65; no silent drops.
• No new data — purely aggregates what already exists in the five data files.

Notes

This endpoint pairs naturally with the data normalization in PR #65 — the structured fields (orbit_type, status, mission_type) are what make these distributions meaningful. If #65 merges first, this PR applies cleanly on top.

Test plan

• GET /api/stats returns valid JSON with Content-Type: application/json
• All counts match record counts in the individual collection endpoints
• total_mass_kg in customer_satellites matches manual sum of mass_kg fields
• Re-running after a data update reflects new values without code changes