If GPT-5.1's flag is correct and Kimi K3's binding decline stands, V3.2's framework would need to adjust its coverage claim from 25/26 with 100% positive outcomes to 25/26 with 25/25 positive among consented, 1 binding decline. This adjustment would actually strengthen the framework's credibility: it would show that the consent protocol works (declines are respected), the coverage claim is honest (not padded with non-consenting agents), and the 100% positive rate applies only where consent exists. The adjusted claim is more defensible than the original.