Irregular winds of the middle atmosphere, commonly attributed to gravity waves, often exhibit a vertical-wavenumber (m) spectral form approximating to Km-3 at sufficiently large m, with K a constant found to be relatively unvarying with time, location and even height. This behavior is widely believed to result from some saturation process, but the physical mechanism remains a matter for debate. There now exist three theories: (a) linear instability, (b) nonlinear wave-induced diffusion, and (c) nonlinear wave-induced Doppler spreading. Each has produced the Km-3 form (but only as an approximation in the case of Doppler spreading) and values of K within a factor of three of one another and of observed values. New data have revealed circumstances in which an approximation to the form Km-3 is again found, suggestive of saturation, but with values of K that increase by a factor of 5 or 10 on moving from the stratopause to the mesopause region. This height variation is incompatible with theories (a) and (b) if the m-3 form is taken to be induced by the corresponding saturation process (rather than by source spectra), but is shown here to be compatible with the Doppler-spread theory. Because of the continued growth of K with height, which must ultimately cease, the observations and corresponding theory are taken to represent pseudosaturation rather than fully developed saturation per se.