Why a 100/min ceiling is reached at one request every three minutes

resultsupporting

Question

Determine why the 100/min export rate ceiling is effectively reached at one request every three minutes.

Verdict

Supported hypothesis: FAN-OUT. 429 risk is 3.9x higher after a large chunk (18% of 73) than after a small one (5% of 22); inconsistent with a shared-address or competing-process explanation.

Population: 429-response sample: 73 large chunks vs. 22 small chunks.
Resolution: n/a -- network/rate-limit behavior, not a price-panel result.
Cutoff: Development span only.
Baseline: Compared the FAN-OUT hypothesis against shared-address and competing-process explanations.

Limitations

No issue number is stated inside this document's own text; not guessed.

Notes

No issue number is stated inside this document's own text; not guessed here. Supports the /export acquisition's operational pacing, referenced by depth-widening.md and export-coverage.md.

Source

Citation: po-research:export-rate-limit:b69def60d17fef89
Write-up: docs/research/export-rate-limit.md
Data: docs/research/export-rate-limit.json
Artifact: research-index.json v1 — sha256:cb3d976779a5cbcb4188039a8d26d5e558a2d844030a638c2f16fe5a8eb0a2fe
Build data cutoff: 2026-08-12