Toole is basically correct but as far as I know he never investigates the small corrections (about +/-1 dB) that are required at certain frequencies in terms of what listeners prefer.
Sorry for the delay in responding. Listener preferences from the earliest double-blind evaluations favored loudspeakers without resonances and flattish axial frequency responses. The next research was to determine the audible thresholds for resonances so that design engineers could make intelligent compromises in lower cost products. Section 4.6 in the 3rd edition describes audible thresholds for spectral tilts and resonances of various Qs. We can detect low Q resonances with measured amplitude deviations of much less than 1 dB when listening to revealing program or pink noise. Toole, F. E. and Olive, S.E. (1988). “The modification of timbre by resonances: perception and measurement”, J. Audio Eng. Soc., 36, pp. 122-142.
So, if listeners prefer neutral - non-resonant - sound these are the ultimate criteria to be applied. Naturally, they can only be addressed using anechoic data.