Top of my table: lasers/EML devices #1 (very high conviction — "most direct optical bottleneck" for 800G/1.6T and co-packaged ramps; weak light-source supply constrains the module ramp). Optical signal processors #2 (power-performance gate). InP substrate supply & epitaxial capacity #3-4 (material every laser starts from). Precision optical packaging #5 (manufacturing-yield gate). Ranks are dated hypotheses from April 2026, each carrying a to-be-validated flag until primary evidence promotes it.