Conduct an independent deep research and adversarial design review for Vercy model WM-ORG-017 Performance Objective / Review, within an enterprise model library that must serve startups and multinational companies. This is a generic public model; no real employee data. Deliver a substantive English research memo, not a marketing summary. Research primary sources on performance management, HR Open Standards, ESCO/SFIA, SPACE developer productivity, evidence quality, corrections and fair review. Distinguish verified links from inferred design. Do not claim compliance or invent clauses. Model boundary: objectives, evidence and attribution, review periods, reviewer assignments, calibration, appeal/correction, recognition, temporal validity and purpose-specific disclosure. Person, employment, organization membership and assignments are separate referenced identities. Rating scales are versioned and contextual; no global intrinsic employee score. No automatic employment decisions from commits, keystrokes or surveillance proxies. Vercy describes a whole object via bundles/layers/findings/questions/data/artifacts and five interacting facets: structure, functions, composition, properties, governance. One model may reference neighbors but must not duplicate their identity. Please provide: (1) exact identity and lifecycle boundaries and alternative model splits; (2) 6-8 bundles, 12-16 layers and ~20-25 concrete findings/questions with evidence artifacts and invariants; (3) relations/cardinalities/time/source-of-truth and permission rules; (4) startup vs international enterprise profiles; (5) a worked synthetic case with two concurrent employments, disputed credit, conflicting evidence, appeal and correction, and change of employer; (6) strongest negative cases and unresolved issues; (7) 8-12 primary source URLs and their actual applicability. Conclude whether usable as a visibly reviewable draft and what specifically blocks canonical assurance. Be independent: no other AI answer is supplied.