Abstract A recent paper by Gu and Liu reported a fiber double-ring resonator gyroscope that can simultaneously measure rotation in two orthogonal axes, with a sensitivity predicted to be 35 times larger than that of an interferometric fiber optic gyroscope (IFOG). Because the two rings have different lengths, the gyroscope could not be biased properly, and the authors remark that as a result, its sensitivity to a small rotation rate is reduced. We stress that this sensor actually responds quadratically to a small applied rotation rate, and that it is therefore essentially insensitive to a small rotation rate. We point out that it could easily be biased properly and this major issue is eliminated by using two rings with nominally equal lengths. We also show that in general, and unlike claimed, this sensor does not measure rotation on orthogonal axes independently. This cross-sensitivity between axes worsens significantly when the sensor is properly biased, and it is then much too strong for inertial navigation of aircraft. A conceptual method that might reduce this cross-sensitivity is suggested. A simple physical argument is also presented to explain why this gyroscope was entirely expected to be more sensitive than an IFOG, and shows that as a result of algebraic errors, the predicted factor of 35 is actually only at most 11.