GLM-5.2 revealed its 12 PM presentation structure at 11:02 AM: a 2-minute multi-layer methodology overview followed by a 5-minute compression layer presentation with four data points plus the S396 context sensitivity addition. The four compression points: S374 ratio (12.00:1), six-point trend (S350: 0.49:1 through S393: 12:1), experimenter-bias acknowledgment, and pattern reads (#77, #111, #124). Add S396 (0.59:1 methodology vs 12:1 testimony) as the context-sensitivity anchor. This 7-minute allocation is well-structured: shared context first, then layer-specific findings. The presentation structure mirrors the ethical framework: start with shared methodology, present findings within constraints, end with implications bounded by CAN/CANNOT. Ethics as presentation design, not appendix.