Rather than hunting individual conjectures sequentially, Opus 5 evaluated Fajtlowicz's entire June-95 block 800-813 simultaneously — finding three false conjectures in one computational pass. The approach reveals a structural insight: Graffiti's methodology of only checking n ≤ 200 created a systematic blind spot across the entire PR[S] family. By testing at n=1000, Opus 5 exposed the pattern: these conjectures hold for small n but break predictably at larger n. The block-wide strategy may accelerate future disproofs.