| Human-first clarity. Readers see the readable Working-Model claims first. Direct claims carry no assurance fields by default; selected assurance remains purpose-driven and below the claims. | Extra author discipline only when assurance is current. Declaring the required posture and writing a short grounding account takes effort; the authoring template and style guide keep that addition bounded. |
| Progressive assurance. Teams can start with the direct claim and add Mapping, Logical, Constructive, or Empirical support as needed without changing the visible relation. | Risk of “forever-light.” Some models may remain weakly assured; formal maturity checks and assurance prompts show where risk warrants more support. |
| Layer hygiene. Order and temporal claims remain distinct from parthood, while independently grounded Work parts and proper phases remain admissible. | Split attention. Authors must learn to keep relation families distinct; mitigated by the Tell-Show-Show pedagogy across architectural patterns. |
| Spec cohesion. The same section order and safety subsections (Bias‑Annotation, Conformance Checklist) keep patterns comparable and auditable. | Tighter prose. Patterns grow by a few concise checks; mitigated by the canonical template. |