Remove Jiangshui screening workflow node while retaining the tested fermentation effect #1519

Record:kb/communities/Jiangshui_LAB_Directed_Fermentation_SynCom.yaml,000356.

The3-node2-arrow graph starts with researcher screening/ratio assembly. Primary abstract supports a three-strain2:3:1 formulation and higher total phenol/flavonoid endpoints versus natural Jiangshui. Remove procedural selection node and arrow; retain the formulation-to-endpoint intervention. Do not turn it into ratio-only mediation, identified compound synthesis, or clinical safety. Preserve all three separately named canonical strains even though each currently has NCBITaxon2.

Complete cached abstract and indexed publisher fragments read; direct ScienceDirect requests403, no full primary body inspected. Fragments do not resolve selected strain/species mapping or fermentation protocol. Preserve the existing anchored taxonomy/protocol knowledge gap and prior history including closed#824. No new structure; keep repair open until protected-main merge.

Dedup:822 all-state issue titles/bodies and14 matching ignored/hidden local files. Closed#824 repaired Jiangshui wording, not these scientific findings. Relevant local contexts inspected; not every historical comment/review. No original figure pixels, external supplements, raw data or code audited. Self-review is not independent approval.