This paper reports a precision audit of a production filter stack against a 13-day window of post-rejection forward-market observations on Solana DEX trading (2026-04-10 to 2026-04-23, UTC). The audit yielded 99,510 follow-up samples across 2,402 unique rejection events spanning eight active filter rules. We classify each event under a five-tier outcome rule and report per-filter distributions. The headline result is the conservative save-to-miss ratio of 3.7 : 1 from windowed measured-drawdown saves alone; every active filter with adequate sample size is individually net-positive. A wider interpretation that credits single-sample-within-60-minute events as saves yields 14.8 : 1. We then test the interpretive premise of that wider tier against the deposited lifecycle data of a separately-published benchmark (RED-2400). The matched comparison shows that early-death-classified mints reach the gone state at 48.9 percent. Non-early-death rejected mints reach gone at 57.6 percent. The early-death classification does not identify tokens at elevated rug-pull risk relative to other rejected tokens. The wider 14.8 : 1 ratio therefore rests on a tier that the matched test does not validate. The conservative 3.7 : 1 is the report the lifecycle data supports. The methodological contribution is the separation of the two evidence bases and the demonstration that the wider tier does not survive matched-comparison testing.
Arati Uday Kamat (2026) studied this question.