Reconcile failed-request evidence across independent collection lanes

For request-scoped canaries, define the failure population as the union of independent evidence lanes such as server outcomes, diagnostic attribution, and retained exact inputs. Each lane can fail independently, so a storage count or database count alone can understate requests. Record lane intersections and collection start times before sampling causes.

Applicability and limits: a deadline is a time-budget result, not proof of unreadability. Inspect full input and derived crops before naming a visual cause. Default-initialized telemetry counters cannot prove that no strategy ran, and server-only events without request or environment identifiers cannot diagnose one input reliably.