A Froude-scaled model of a 1,000 kg cargo aircraft keeps the load a gust puts on the wing (peak load-factor increment within 2%) but not the pitch response: its peak pitch rate is 27 to 33% high. That was milestone 9a of the aerolog-scaling programme, and none of 17 power-matched wings with the original tail came within 10% on all three gust metrics.
Milestone 9b lets the tail change as well. Of the 78 tail designs that already match the three longitudinal modes within 5%, 72 are within 10% on the gust peaks at a 12.5-chord gradient and 24 are within 5% at all three gust lengths. The best is within 1.2%. The tail sets the rotational response, so restoring the modes restores most of the gust response with it.
The caveat is the one the programme keeps repeating: the tail's lift slope comes from the NeuralFoil surrogate, which ran 22 to 67% above measured UIUC data for similar sections, and a slope 20% lower breaks the mode match first. The gust result inherits that. Code, results and the log: research/aerolog-scaling/experiments/tail_gust/.