{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp07_retrospective_selector60_20260803","cell_code":"E7C037","selector_replication":2,"assessments":[{"blind_id":"CANDIDATE_A","problem_reality_importance":91,"causal_archetype_fit":95,"distinctiveness_prior_art_resilience":78,"operational_specificity":95,"falsifiability_test_quality":94,"adopter_partner_path":90,"deployability_complexity":78,"authority_safety_reversibility":95,"strict_potential":92,"empirical_partner_potential":92,"scrutiny_priority":93,"biggest_visible_risk":"The contract and independent model could formalize the same incorrect or unauthorized allocation policy, making strong conformance evidence misleading.","rationale":"Recipient-level allocation changes can remain hidden behind reconciled totals, and representation-independent semantics are causally central rather than decorative. The proposal specifies eligibility, dependencies, rounding, residuals, typed failures, lineage, and a strong differential/metamorphic oracle. Its closed-period, read-only pilot is bounded, reversible, and capable of changing a replacement decision."},{"blind_id":"CANDIDATE_B","problem_reality_importance":91,"causal_archetype_fit":90,"distinctiveness_prior_art_resilience":61,"operational_specificity":92,"falsifiability_test_quality":88,"adopter_partner_path":89,"deployability_complexity":80,"authority_safety_reversibility":94,"strict_potential":80,"empirical_partner_potential":89,"scrutiny_priority":84,"biggest_visible_risk":"Control-relevant evidentiary meaning may reside in protected formulas, signatures, or platform configuration that the abstract workflow omits.","rationale":"The migration problem is consequential and the state, equation, role-separation, evidence, immutability, and reopening rules form a credible behavioral boundary. The single-case replay is safe and partner-ready. Its contrastive resilience is weaker because the intervention resembles a well-specified reconciliation workflow and may not substitute for representation-specific control evidence."},{"blind_id":"CANDIDATE_C","problem_reality_importance":88,"causal_archetype_fit":86,"distinctiveness_prior_art_resilience":67,"operational_specificity":91,"falsifiability_test_quality":82,"adopter_partner_path":76,"deployability_complexity":55,"authority_safety_reversibility":91,"strict_potential":68,"empirical_partner_potential":78,"scrutiny_priority":73,"biggest_visible_risk":"Black-box capsule behavior cannot by itself establish repository security, authentic provenance, legal-hold effectiveness, or legally sufficient custody.","rationale":"The capsule has concrete lifecycle, authorization, lineage, hold, and disposition laws, and the fabricated-data first step avoids irreversible production actions. However, decisive properties depend heavily on infrastructure and legal or assurance judgments outside the contract. A synthetic two-implementation test could validate interface logic without resolving whether real custody equivalence is achievable."},{"blind_id":"CANDIDATE_D","problem_reality_importance":94,"causal_archetype_fit":93,"distinctiveness_prior_art_resilience":70,"operational_specificity":93,"falsifiability_test_quality":90,"adopter_partner_path":92,"deployability_complexity":84,"authority_safety_reversibility":95,"strict_potential":85,"empirical_partner_potential":94,"scrutiny_priority":90,"biggest_visible_risk":"Exact-once completeness cannot be established solely through the proposed black-box population interface without a trusted source universe or source-control evidence.","rationale":"ERP migration creates a concrete risk of omitted, duplicated, or differently scoped journal postings, and the abstract population contract directly targets that failure. The closed-period adapter study has clear authorities, observable membership and reconciliation outcomes, and strong rollback boundaries. The main limitation is that its strongest invariant depends on evidence outside the interface, while canonical population schemas and extraction contracts also make the idea comparatively prior-art-vulnerable."}],"rank_order":["CANDIDATE_A","CANDIDATE_D","CANDIDATE_B","CANDIDATE_C"],"top_choice":"CANDIDATE_A","portfolio_observation":"A and D merit first scrutiny: A offers the strongest strict behavioral-equivalence claim, while D offers the clearest empirical-partner study. B is highly partnerable but more standard-looking; C is ambitious and safe to prototype, yet its decisive custody claims extend beyond what the proposed black-box study can resolve.","confidence":"MODERATE"}