Consistency: Consistent on the core V-ATPase regulatory-subunit role. PN treats H purely as a "V1 proton-pump component"; the review adds a well-supported SECOND function (AP2M1/AP-2 binding → clathrin-mediated endocytosis, GO:0006897 / GO:0035615 clathrin-cargo adaptor; PMID:12032142) that the PN node does not represent. Not a contradiction, but the PN row under-describes H.
PN story / NEW pressure: Projected terms VERIFIED real. GO:0046612 (lysosomal V1) is a lysosome-specific sibling of the review's vacuolar GO:0000221 — already captured at function level. GO:0007042 lysosomal lumen acidification IS already in the review (IBA + 2× NAS, ACCEPT) → "already_in_goa_exact," correctly. GO:0033176 already present (ACCEPT/entailed). No NEW pressure from the PN node; if anything the gene's endocytic-adaptor function is the under-captured story, and that is project-orthogonal (not part of the ALP V-ATPase row).
Mapping strategy: No change to node mapping warranted from this gene. Projected CC terms are at/below review specificity; no broader-than-review over-reach.
Evidence alignment: As with G1, PN cites review-article titles (mTORC1/V-ATPase/neurodegeneration) absent from the review's PMID set; review is anchored on PMID:9442887 (Forgac review), PMID:33065002 (cryo-EM), PMID:12032142 (AP2M1). Same V-ATPase biology; the endocytic-adaptor literature is unique to the review.
Verdict: Consistent, already captured (all three projected terms present or covered by siblings). No edits needed for the PN row; note that H's AP-2/endocytosis function lies outside the PN node and is fully handled in the review.