Consistency: Deep research ↔ review ↔ PN annotation consistent. VIPL is an L-type lectin, NOT a glycosidase — matching the "Lectin chaperone" type. Distinctive vs its cycling paralogs: it is a non-cycling ER-resident (RKR retention) regulator of ER export / regulator of ERGIC-53 (core BP GO:0006888, GO:0015031). D-mannose binding accepted as a homology/CRD-based MF (no direct binding assay for VIPL itself — review states this clearly). No contradictions.
PN story / NEW pressure: No NEW GO pressure. ER-export-receptor / ERGIC-53-regulator role is captured by transport terms; a candidate physiological client (ITGA6, via PMID:38687323) is noted but single-paper/candidate-stage and the review correctly did not mint a term. Conclude: already captured.
Mapping strategy:Group→GO:0006487 (new_to_goa) over-reaches as for the other lectins — VIPL reads/exports high-mannose glycoproteins; it does not perform N-linked glycosylation. Additionally, VIPL is predominantly ER-resident (not a classic ERGIC cycler), so even the family-level "N-glycosylation system" framing fits it least well. PN node correct to leave "Lectin chaperone" unmapped; projected term should not propagate.
Evidence alignment: PN dossier is mapping-only. Review evidence centers on PMID:12609988 and PMID:12878160 (PubMed-verified primary papers) plus disease PMIDs (26566883, 31020005, 37667433) and HCMV/TRC8 (38687323). No bibliographic conflict; divergence is conceptual (group biosynthesis term vs. lectin/ER-export function).
Verdict: Review strong and PN-consistent (ER-resident lectin/export regulator, not a mannosidase). Inherited group→GO:0006487 projection is an over-reach. Recommended edits: [MAP] Do not propagate GO:0006487 (protein N-linked glycosylation) to LMAN2L/VIPL — it is an ER-resident high-mannose lectin / ER-export regulator, not an N-glycosylation enzyme; flag the "N-glycosylation system" group node as over-broad (and a poor fit for a non-cycling ER resident).