Exact results for the per-bit energy floor, the invasion front and osmotic minimum work — proofs for questions 16, 10 and 17 of the notebook's twenty promising science questions.
| # | Question | Verdict | Exact result |
|---|---|---|---|
| 16 | How closely can a link approach its signal-energy floor? | Confirmed in part | Eb/N0 > ln 2 = −1.59 dB, approached monotonically; no finite optimum |
| 10 | Which growth and dispersal combination sets invasion speed? | Refuted | v = 2√(rD), a geometric mean; the ratio r/D predicts nothing (R² = 0.08) |
| 17 | How does water recovery change ideal separation work? | Direction confirmed | w = πf[−ln(1−r)]/r, strictly convex; the claimed πf/r crosses the true curve at r = 1 − 1/e |
Not peer reviewed. No theorem here is new — the results are Fisher (1937), Kolmogorov–Petrovskii–Piskunov (1937), Shannon (1948), Bramson (1983) and Wang–Violet–DuChanois–Elimelech (2020). What is new is the comparison, and the independent numerical replication in code/. Reproduce with cd code && ./run_all.sh.