Use Hyde 2022, PDF sheets 187–203, for the located book method; the linked Ready citations to Lost Lantern, Callaway, Fraley, and VinePair provide independent professional and consumer-practice corroboration.
A proposed blend should be tested on labeled small samples using controlled dilution, measured ratios, comparative tasting, and written notes before the formula is applied to the full batch.
Hyde's located procedure is independently corroborated by Lost Lantern, Dan Callaway, Nancy Fraley, and VinePair evidence covering iterative prototypes, scale-up, high-impact component control, and separate bench testing. All Step 1 web paraphrases and citation records were publication-checked on 2026-08-29.
Bench success does not guarantee identical behavior after marrying, proofing, filtration, resting, or full-scale mixing. Sensory bias and measurement error still require controls.
A small measured trial protects the base stock from irreversible changes, permits side-by-side comparison, and produces a formula that can be repeated or scaled.
Publication-stage validation completed 2026-08-29. Evidence is cross-source, locators are confirmed, citation-ledger records are Ready, and no direct quotation is stored.
Basis
Hyde gives a located, stepwise practitioner method built around labels, controlled dilution, measured ratios, notes, comparison, and delayed scale-up.
Current judgment
The method is Supported and independently corroborated across professional, producer, and consumer-practice sources covering iterative prototypes, scale-up, high-impact component control, and separate bench testing.
Drafting guardrail
Present bench blending as risk reduction and experimental discipline, not as a guarantee that the scaled or rested blend will behave identically.
Hyde complete-copy audit
The methodological claim and other source evidence are retained. Hyde's201–202 numeric example contains a ratio/total error:600ml at8/60/32% requires48/360/192ml, not48/300/192. Equal water additions190 do not standardize unequal ABVs. Cite the trial method, not the faulty calculation; independently check arithmetic, contraction, temperature and final proof.