A10-a/Q4 · MSA: UI and exporter label platform ndc bands as "AIAG Acceptance Criteria"

[Siddardth7]

Split out of #223 (audit A10-a) by SME decision — deferred because it churns UI-string tests and its wording depends on #223's Q2 outcome.

Problem

The ndc sub-bands 2 ≤ ndc < 5 (Marginal) and ndc < 2 (Reject) are not AIAG. Verified against the on-machine manual during #223's citation sweep: AIAG states only

This value should be greater than or equal to 5.

(MSA_Reference_Manual_4th_Edition.md lines 2214 and 3427). The 2–4 and < 2 sub-bands appear nowhere in the manual.

Two user-facing surfaces attribute them to AIAG anyway:

• apps/msa/msa_app/pages/gage_study.py:168 — renders the subheader "AIAG Acceptance Criteria" over a block containing both bands.
• apps/msa/msa_app/exporter.py:58-70 (VERDICT_SENTENCES) — repeats them, without an AIAG label.

This is the same defect class as #223's fabricated quotations: an unearned appeal to primary-source authority. It is an attribution, not a quotation, which is why it fell outside #223's strict acceptance criterion.

Dependency

#223 Q2 decides the wording. That was resolved as: relabel the bands in ASSUMPTIONS_LOG.md as an internal design choice built on AIAG's single ndc ≥ 5 rule. This issue should use consistent language on the UI and exporter surfaces.

Scope

• Retitle the gage_study.py subheader so the AIAG-sourced %GRR bands and the platform-defined ndc bands are visibly distinguished (e.g. "Acceptance Criteria (AIAG %GRR bands + platform ndc bands)")
• Give the exporter's verdict sentences the same distinction where they imply a standard
• Update tests/test_pages_gage_study.py and any exporter string assertions

Constraint

No logic, threshold or verdict-mapping change. The bands themselves stay exactly as they are; only their attribution changes. If a band's value looks wrong, that is #224's territory, not this issue's.