docs(seed): clarify passenger-km and round-trip in business travel transport

[mrivas00]

Summary

• A user reported that the help panel of Viajes de negocios — Traslado left two things unanswered: whether the quantity is one person's kilometres or people × kilometres, and whether a trip counts one leg or both. Neither question was addressed anywhere on the page.
• The answer is not uniform, which is why the page could not be fixed with a one-liner: the DEFRA 2025 factors seeded for this subcategory use two different bases — air, bus and rail are per passenger-km, while car and taxi are per vehicle-km. The page now states that distinction up front, says both legs of a round trip count, and gives the formula that builds the quantity (km per leg × legs × people × trips, dropping the people term for car/taxi).
• While verifying the numbers against tools/seed/src/data/base/methodologies.json, the page turned out to contradict the seeded data in several places, so it is now aligned with what the calculator actually offers.

Linked issues

None — reported as direct user feedback.

Type of change

• feat — new feature
• fix — bug fix
• refactor — internal change with no behavioural impact
• docs — documentation only
• chore — repository or tooling change
• infra — infrastructure-as-code change
• security — security-relevant change (also fill in the security
  checklist below)
• breaking — backward-incompatible change for country deployers
  (must include a deprecation path or migration note)

What was contradicting the seed

Page said (before)	Seed says
Factor table with invented values (0,25 / 0,15 / 0,03 / 0,18 kg CO₂e/km)	The 10 real Transporte options with their DEFRA 2025 values (0,1257 / 0,117 / 0,1091 / 0,1038 / 0,0354 / 0,148 / 0,173 …)
Apply a cabin-class multiplier by hand (×1,5 / ×2-3 / ×3-4)	Class is part of the option itself (… Economy / … Business), so no manual multiplier — Business is already ~1,5× (short haul) to ~2,9× (medium/long haul)
Flight bands by distance (<500 km / 500-3.700 / >3.700)	Bands by duration (Short haul (<3hrs) / Medium haul (3-6hrs) / Long haul (>6hrs))
Unidad: km, moneda local	allowedMeasurementUnitsAbbreviations: m, km, mi — no currency unit exists
"Si vas con colegas en el mismo Uber o taxi, divide los km entre los pasajeros"	Taxi factor is per vehicle-km, so dividing leads to under-reporting; the vehicle's kilometres are counted once

Country-agnosticism checklist

• No country-specific logic was hard-coded in the codebase.
• Country variation is expressed through configuration, seed data
  or system parameters.
• The change is backward-compatible with existing country
  deployments, or a deprecation path and migration note are
  documented in this PR.

DPG checklist

• Documentation under docs/ was updated where relevant.
• No new platform-specific dependency was added without an
  open-source alternative path documented.
• No new PII handling pattern was introduced without review
  against docs/security/sensitive-data.md and PRIVACY.md.
• If this PR adds user-facing strings, they are in Spanish (the
  project's user-facing language).

Security checklist

• N/A — this PR does not touch security-sensitive code.

Mandatory local checks

pnpm format:check   # pass
pnpm lint           # pass (13/13 tasks)
pnpm type-check     # pass (13/13 tasks)
pnpm test           # not run — see reviewer notes

• All of the above pass locally on my machine.

Screenshots / evidence

The rewritten worked example, recomputed against the seeded factors:

Transporte	Quantity	Factor	Emissions
Avión Long haul Economy (3 people, round trip, 9.500 km/leg)	57.000 km	0,1091	6.219 kg CO₂e
Avión Medium haul Economy (4 trips, 1 person, 2.100 km/leg)	16.800 km	0,1170	1.966 kg CO₂e
Avión Short haul Economy (3 trips, 2 people, 620 km/leg)	7.440 km	0,1257	935 kg CO₂e
Bus (5 people, round trip, 225 km/leg)	2.250 km	0,1038	234 kg CO₂e
Taxi (vehicle kilometres, occupants not multiplied)	600 km	0,1480	89 kg CO₂e
Auto (vehicle kilometres, occupants not multiplied)	800 km	0,1730	138 kg CO₂e

Total ≈ 9.581 kg CO₂e (~9,6 t). The example closes by explaining the two effects that caused the original confusion: the international conference is 65% of the total because of passenger-km (3 people × 2 legs), not because its factor is high — it is in fact the lowest air factor — and shared taxis contribute little precisely because occupants are not multiplied.

Reviewer notes

• The page is only reflected in the app after reseeding. It is loaded by tools/seed/src/scripts/seedExplanations.ts, so an existing deployment keeps serving the stored copy until explanations are seeded again.
• pnpm test was not run: the diff is a single seed markdown file with no code path, and no test asserts on explanation content. Happy to run the suite if you would rather have it on the record.
• Worth a second opinion on two judgement calls: the km-equivalence hint I added for choosing a flight band (short ≈ up to ~2.000 km, medium ≈ 2.000-5.000, long ≈ >5.000 — marked as orientative, since the seeded options are defined by duration), and the RFI note, which now says the platform applies the DEFRA 2025 factor as-is and that any radiative-forcing adjustment belongs in "Factor propio" rather than in the quantity.
• The ferry bullet is kept in the coverage list even though no ferry option exists in the dimension, so the scope of the subcategory is not silently narrowed.
• Two red checks are pre-existing and unrelated to this diff (a single markdown file): Audit and Trivy Image Scan (api) both fail on the same transitive dependency, deepmerge-ts 7.1.5 pulled in by @prisma/client → prisma → @prisma/config (CVE-2026-40345 / GHSA-ggr8-5vv4-36mx). Reproduced locally on this branch, and already tracked in fix(deps): force deepmerge-ts to 8.x to patch CVE-2026-40345 #602. Everything else passes, including Seed Unit Tests, Format Check, Lint, Type Check and the full test matrix.