The Lipzhitz et al. paper "
An Investigation of Sound Radiation by Loudspeaker Cabinets" was written in 1991 -- 3 years after Floyd Toole published our paper on the detection of resonances, "
The Modification of Timbre by Resonance: Perception and Measurement"
In re-reading the Lipshitz paper they conclude that some of the cabinet resonances on the worst-case speakers may be audible or "borderline audible" on certain signals based on our predictions of audibility of resonances.
The audibility of a resonance depends on its amplitude, Q, frequency and the signal exciting it (steady-state versus transient). The audibility of a resonance at a given level increases as the resonant frequency increases, and the Q decreases. So shifting the resonance up in frequency and lowering its Q will increase its audibility -- all things being equal.