Eehm, then what are these 22µF X5R types doing on my desk?

But you wouldn't use X5R types in a high quality audio application, would you?
Several years ago I had measured harmonic distortion of several high capacitance ceramic caps. If the time constant (R x C) of single pole high pass filter was choosen standardly for common audio signals, harmonic distortion at lower frequencies rised up dramatically. Only if time constant was choosen very high, means if capacitance was rised to abnormal values, harmonic distortion was acceptable (<0.01%).
An aluminium electrolytic isn't much more linear either, of course, but it's muche easier to provide this 'abnormal' capacitance. Aluminium electrolytics are put in the signal path of studio recording mixing consoles to hundreds over hundreds and nevertheless over all harmonic distortion is amazingly low.

Another property which speaks against the use of high capacitance ceramics in high quality audio applications is the extreme high 'Er', which results in lots of dielectric absorption. In many publishings dielectric absorption is stated to be the major cause of signal degradation.

Kai