docs: define sustained top-five ownership vocabulary

[svanlabs]

What generated this?

Generated-by: codex/gpt-6

Motivation

The operator settled what sustained all-five leaderboard ownership means. The glossary needs names that distinguish sampled occupancy from an isolated sweep and keep missing observations visible.

Closes #958

Solution

Add two terms in the docs layer: sampled sweep occupancy and sustained top-five ownership. The acceptance contract remains canonical in Choose measurable criteria for sustained ownership of places 1–5; the glossary stores meanings only. This can merge independently of the existing research/glossary PR 961.

Validation

Full gate on commit 882e1bd: 885 passed, 0 failed, 4 ignored across 35 binaries; formatting, clippy, dependency/notices checks, tool suites, documentation checks and both TypeScript configurations passed. No poker or performance behavior changes.

== placeholder markers
== ai provenance
provenance: 37 commit(s) in HEAD name their system
provenance: AI-PROVENANCE.md lists the 9 system(s) in HEAD
== rustfmt
== Cargo.lock matches the manifests
== file sizes (docs/CONTRIBUTING.md: 500-line limit, baseline may only shrink)
== clippy -D warnings
== cargo-deny (licenses, bans, sources)
== cargo-deny advisories
   (cargo-deny advisories: 2 s)
== third-party notices current
== tickets lint + tool tests (tickets, codec vs zlib, test runner, release/rollback, keepalive, update, adopt-upstream, file-size, build lock, systemd units)
   (tickets lint + tool tests (tickets, codec vs zlib, test runner, release/rollback, keepalive, update, adopt-upstream, file-size, build lock, systemd units): 3 s)
== docs name only paths and commands that exist
   (docs name only paths and commands that exist: 101 s)
== workspace tests
test.py: 885 passed, 0 failed, 4 ignored in 35 binaries; build 0.2s, run 4.5s (6 jobs x 2 threads, profile gate, peak test RSS 195 MB)
   (workspace tests: 7 s)
== tsc
== ok (full, 118 s)

Checklist

• scripts/check.sh full passes on this exact commit; output is above.
• One focused documentation change.
• Behavior simulation and golden changes are not applicable: glossary only.
• Performance measurement is not applicable: glossary only.
• Documentation paths and commands pass the drift check.
• Generated-by: is in the commit footer and this body.

Validation refreshed at 10db5fbe1c490f791acadc5dcdfe49464234a0c8 against main 15c3d730814f5e9465b6dc0b34073052b071fee3: shared local review runner/full gate passed, 892 passed, 0 failed, 4 ignored, 35 binaries. Independent Standards and Spec reviews found no substantive issues.

Generated-by: codex/gpt-6

[svanlabs]

Reviewed integration head 10db5fbe1c490f791acadc5dcdfe49464234a0c8 against 15c3d730814f5e9465b6dc0b34073052b071fee3 using the shared local runner and independent reviewers.

Standards

No substantive violations found. The two glossary definitions retain missing/stale observations in the scheduled denominator and limit claims between samples. Numeric thresholds remain in the canonical decision record, avoiding duplicated acceptance contracts. No runtime behavior or promotion threshold changes.

Spec

No findings. Both settled terms required by #958 are present. Sampled sweep occupancy excludes unknowns from success; sustained ownership requires all-five evidence over complete seasons and official final standings. This follows the accepted instruction that glossary entries contain meanings while the decision comment remains the acceptance contract. It does not claim the fleet already passes.

Validation: exact-head full gate passed: 892 passed, 0 failed, 4 ignored; 35 binaries.

Summary: Standards 0 findings; Spec 0 findings.

Generated-by: codex/gpt-6