Changing the UMX Gee Bee's prop to a 5 x 2.75 reveals that AS3X can induce substantial, spurious flight control inputs just like the Blade mCP X helicopter with AS3X.
This development validates and confirms that AS3X is in fact what is inducing a loss of model control:
I first noticed a near total loss of flight control induced by AS3X when my mCP X helicopter would try to careen out of control at certain throttle settings and/or blade loads. But I didn't fully understand that the issue was caused by low frequency structural resonance until I discovered a massive improvement in heli controllability after the addition of the Z8RC boom truss modification. The Z8RC boom truss functions in the harmonic space to suppress standing sine wave resonance.
Now the UMX Gee Bee suffers from the same AS3X-induced loss of flight control in certain resonance bands. This is not typical of other 3-axis gyro systems. It appears that AS3X sensors, gains and/or logic are especially susceptible to low frequency resonance, causing compounding amplitude oscillations and loss of flight control authority. As a result of the multi-model manifestation of this problem, I am issuing the following warning:
WARNING: In certain structural resonance bands, AS3X may induce increasing flight control oscillations. An immediate change in throttle setting is advised to avoid loss of model control.
This warning should accompany further sales of AS3X-enabled models until E-Flite resolves their defective control logic.
