Froude scaling promises that a model keeps trim speed and mode frequencies in scaled units. It does so only if the coefficients stay fixed, and at 1:10 the Reynolds number is about 1e5, not 3e6. In aerolog-scaling a 4-state longitudinal model with NeuralFoil polars (a surrogate for XFOIL) shows what that does for a tripped wing: flown at the full-size trim angle, the model trims 6.5% faster and its short-period frequency moves 5.1%, so the 5% bar in hypothesis H1 fails. Flown at the full-size lift coefficient, frequencies stay within 1.3% but short-period damping drops 6.3%.
Free-transition runs at 1:10 enter a laminar-bubble regime where the surrogate is not trusted, and phugoid damping depends on a thrust law the model lacks, so neither is reported. There is no gust response yet, and the polars are not checked against measured UIUC data.