fix(overton): address 6 substantive audit findings across all report surfaces

- NSC sensitivity: only 3.1% of CS surge attributable to NSC inclusion
- Submitter parsing: opposition-only d=0.85 likely inflated (~0.65-0.75)
- Mechanism taxonomy: relabeled as exploratory (kappa=0.41, 50.5% agreement)
- Causal claims: softened from 'refuted' to 'less consistent with data'
- 2026 reversion: material moderation persisted but shock was primary driver
- Placeholder contamination: r=0.43 drops to r=0.34 without 6K (1,1) defaults

reports/overton_window/overton_report.html

@@ -137,7 +137,7 @@
   <!-- 3. Methodology -->
   <h2>Methodology</h2>
   <p>Each motion was scored on two independent dimensions by an LLM-based classifier: <strong>stijl</strong> (rhetorical extremity, 1&ndash;5) and <strong>materieel</strong> (policy impact, 1&ndash;5). The gravity level is derived from the combined scores. Centrist support is measured as the fraction of centrist parties that voted in favour. The <strong>canonical definition</strong> uses 4 strict centrist parties: D66, CDA, CU, and NSC. A broader <strong>all-party</strong> model (including VVD, BBB, and other non-extreme parties) is shown for comparison.</p>
-  <p>The dataset contains <strong>29,591</strong> motions. The pre-2024 period (January 2016 to June 2024) covers 21,695 motions; the post-2024 period (July 2024 to present) covers 7,875 motions. Mean stylistic extremity: <strong>1.36</strong> (on a 1&ndash;5 scale). Mean material extremity: <strong>2.12</strong>. The Pearson correlation between the two dimensions is <strong>r = 0.43</strong>, confirming that style and substance are separable but moderately related.</p>
+  <p>The dataset contains <strong>29,591</strong> motions. The pre-2024 period (January 2016 to June 2024) covers 21,695 motions; the post-2024 period (July 2024 to present) covers 7,875 motions. Mean stylistic extremity: <strong>1.36</strong> (on a 1&ndash;5 scale). Mean material extremity: <strong>2.12</strong>. The Pearson correlation between the two dimensions is <strong>r = 0.43</strong> (excluding placeholder motions, r = 0.34), confirming that style and substance are separable but moderately related.</p>
 
   <!-- 4. Gravity-Controlled Analysis -->
   <h2>Gravity-Controlled Analysis</h2>
@@ -537,12 +537,12 @@
   <h2>Key Takeaways</h2>
   <div class="verdict">
     <ul>
-      <li><strong>Overton shift confirmed for right-wing motions.</strong> Centrist support for motions proposed by right-wing parties rose from 0.384 to 0.620 after July 2024, while support for other party motions stayed flat. The shift is real and asymmetric.</li>
+      <li><strong>Exploratory evidence suggests an Overton shift for right-wing motions.</strong> Centrist support for motions proposed by right-wing parties rose from 0.384 to 0.620 after July 2024, while support for other party motions stayed flat. However, inter-annotator agreement on mechanism classification was moderate (κ = 0.41), meaning the mechanism evidence should be treated as suggestive rather than conclusive.</li>
       <li><strong>Migration is the gateway domain.</strong> Asylum/migration is where the Overton shift is most genuine &mdash; centrist support more than doubled (0.153 &rarr; 0.369) while material impact barely declined. Centrists went from zero support for the most extreme migration motions to nearly 20%. This domain is where right-wing parties first perfected the consensus framing and institutional appeals that later spread to climate, security, and economic policy.</li>
       <li><strong>Centrist tolerance increased at all gravity levels.</strong> Under both the all-party and strict 4-party centrist models, post-2024 centrist support was higher at gravity levels M3 through M5. The strict centrist core moved as well, not just the broader coalition.</li>
       <li><strong>Shift is not across the board &mdash; M5 motions remain largely rejected.</strong> Even after the window shift, motions at the highest gravity level (M5) received only 10.1% strict centrist support. Some positions remain firmly outside the Overton window regardless of the political climate.</li>
       <li><strong>Style and substance are only moderately correlated.</strong> With r = 0.43, the rhetorical framing of a motion and its substantive policy impact move partly independently. This means motions can be substantively radical but rhetorically cautious (like motion 144), potentially slipping through the window under the radar.</li>
-      <li><strong>Post-2025 downward trend suggests shift may be reversing.</strong> Strict centrist support for M3+ motions peaked in 2024 (0.487) and declined through 2025 (0.451) and 2026 (0.297). This trajectory raises the question of whether the Overton window shift was a temporary realignment rather than a permanent change.</li>
+      <li><strong>Post-2025 downward trend suggests shift may be reversing.</strong> Strict centrist support for M3+ motions peaked in 2024 (0.487) and declined through 2025 (0.451) and 2026 (0.297). The 2026-Q1 reversion (CS = 0.334) shows material moderation persisted while stijl reverted. Critically, CS was already declining through 2025 despite continued policy moderation, suggesting the 2024 spike was primarily an electoral shock response rather than a sustained realignment.</li>
     </ul>
   </div>