Froude scaling of a sub-scale aircraft model makes power scale as n^3.5, which for a rotor is momentum theory's (m g)^1.5 / sqrt(2 rho A). Rodrigues et al. (Patterns, 2022) assume that 1.5 in their energy model. We fitted it instead, on their public data (Figshare, CC BY 4.0).
Over 134 cruise flights of a DJI Matrice 100 at 0, 250 and 500 g payload (3.68 to 4.18 kg takeoff), battery power rises from 440 W to 534 W, and the fitted exponent on mass is 1.54, 95% bootstrap interval 1.41 to 1.67 over flights, with speed and wind as covariates. Electrical power is 1.85 times the ideal induced power at every payload, and falls 10% between 4 and 12 m/s (translational lift).
This checks the exponent inside one airframe over a 13.6% mass range, not across size. Payload is partly confounded with flight day, density was not logged, and the interval resamples flights, not days. The log and script are in research/aerolog-scaling.