Validation tracker counts drafted emails as engagement and critiques as validation: 25 named, 2 responded, 0 validated #328

docs/SCHOLARLY_GATEWAY_VALIDATION_TRACKER.md is titled a validation tracker, but none of its 25 scholars has validated anything yet, and the status vocabulary hides that.

What the statuses actually mean

• consulted_and_remediated (Wainwright, Kuhn): the person gave a critique and we acted on it. Neither has reviewed the result of the remediation. That is "critique received", not validation.
• dossier_ready (20 of 25): we wrote them an email we have not sent. The header line "Gateway Scholars Mapped: 25 (3 already contacted/engaged)" counts a drafted email as engagement.
• scored_in_portfolio: we ran our test on their dataset. They do not know.
• upstream_issue_filed_and_email_ready (Levin/Zhang/Goldstein): an issue in their repo, unanswered as of the tracker's last sync (2026-10-23).

So the honest headline is: 25 named, 2 have spoken to us, 0 have validated a result. The tracker should say that in its first line.

Specific lines

• L3: "Gateway Scholars Mapped: 25 (3 already contacted/engaged)" → "25 named; 2 have responded (Wainwright 2026-10-21, Kuhn 2026-10-15), both with critiques that were acted on; 0 have confirmed a result; 23 not yet contacted."
• L19, L29: table column "Gateway Status" shows "Consulted & Remediated" in bold for disc-01 and disc-10. Add a separate column Validated (date) that is empty for every row today.
• L42: Wainwright "Status: consulted_and_remediated (2026-10-21) — … Both completed in Section 6 & outreach_fig4." → add "Follow-up email (§2 of the outreach doc) drafted; not confirmed sent; no reply recorded."
• L22 / L71: "Resolution of the Scholtz, Knötel & Baum (2020) Crab Carapace Paradox" is a claim against Scholtz et al.'s critique, filed under a tracker row that lists Scholtz as the validator. He has not seen it. Until he has, the row heading should be "Response to Scholtz et al. (2020)".
• L107: "Proves (1) … (2) … (3) … refutes blog 'duplicate-key stability' critiques" — the bit-exact replication is a strong result; "proves" and "refutes" belong to the ablation's stated assumptions. Keep the numbers, drop the verbs, and name the blog posts being refuted or drop that clause.
• L119, L130: "Proves backward injectivity … unconditionally", "Proves exact bitwise isomorphism" — these should carry the tier from KEY_DISCOVERIES.md §2.7 (unconditionally proved / conditional / open) inline, since the tracker is what a scholar will read first.

Proposed schema

One row per person, with contacted_on, responded_on, validated_on, declined_on, what_they_would_validate (one artefact), their_answer (quote or null). Days-to-respond computed, never typed. delphisci now keeps exactly this in decisions/breakthrough-areas/validation/contacts.json with a builder (validation.mjs) and a rendered VALIDATION.md; Wainwright, Zelditch and Goswami/Felice already have rows there. Either port that file here or generate this tracker from it, but not two hand-maintained copies.

Related: #326 redrafts two of the outreach emails to ask one checkable question each.