Consistency: Consistent. Deep research/notes ↔ review agree: non-catalytic B (brain) subunit of V1, required for assembly, ubiquitous; melanosome and renal apical roles are correctly KEEP_AS_NON_CORE. PN frames B2 generically as a "V1 lysosomal v-ATPase proton pump component," which the review supports (GO:0000221 V1 domain ACCEPT; GO:0007035/1902600 ACCEPT).
PN story / NEW pressure: No over-reach, no unmet pressure. PN's mTORC1-upstream row is generic to all V1 subunits; the B2 review does NOT add mTORC1-specific annotations (unlike A/D/E1), and correctly does not invent them — defensible, since no B2-specific mTORC1 evidence exists. GO:0007042 is absent from B2's GOA (dossier: more_specific_than_existing_goa; confirmed 0 GOA hits) — a defensible ADD. GO:0046612 (verified real) also absent — defensible more-specific ADD.
Mapping strategy: B2 does not change node mapping. Note B2 lacks the lysosomal-specific complex terms (GO:0046611/0033176 absent from its GOA), so the subtype-complex projection of GO:0033176 here is genuinely additive (entailed_by_goa_closure per dossier) rather than redundant — appropriate.
Evidence alignment: Minimal overlap. PN cites generic V-ATPase/mTORC1 review titles; review anchors on B2-specific primaries: PMID:2145275 (brain B-isoform cloning), 12643545 (melanosome), 29993276 (renal), 33065002 (cryo-EM), 32001091 (= one PN review, overlap). PN evidence is non-B2-specific; review evidence is gene-specific and stronger.
Verdict: Consistent / ADD GO:0007042 + GO:0046612 (both verified, both genuinely new to B2 GOA); no contradictions. Recommended edits: none required; mappings sound as-is.