The aim is to acquire low variance roll responses (performance robustness) of control of an Aerosonde despite the high level of tolerances in aerodynamic parameters and working speed. In this respect, fractional-order proportional plus integral and derivative (FOPID) is a valuable option; others are H∞ and μ synthesis. FOPID can tolerate system uncertainty by maintaining a wide open-loop flat phase margin band. All three methods are worked out using the linearized system model and deliver (at least initially) high-integer-order controllers. The uncertainty level is not explicitly considered in H∞, but it may be presented in the μ synthesis and FOPID. The uncertainty presentation in the modified fractional-order controller (mFOC) design is through a Φd curve. The Φd curve is fitted to the mean of the upper and lower bands of the phase margins distribution map of the random systems. It is shown that the mFOC design perfectly secures the desired phase margin flatness. The controllers are applied to the roll of an unmanned aircraft vehicle with a 30% tolerance in the aerodynamic parameters, and operation speed and robustness in performance is evaluated. The simulation results indicate that the mFOC design renders more coherent responses than what H∞ and µ synthesis design deliver. This is confirmed through extensive simulations.