Direct anechoic response is not terribly useful without also off-axis anechoic response.
Mention of "transition frequency" brings into scope a great lot of discussion about if that even exists, if it's primarily psychoacoustic, or if it relates to analysis methods and nonstationarity of data.
There is very very much a transition in acoustics starting at 500Hz, and ending at about 2kHz, based on the maximum firing rate of the inner hair cell, for instance.
There is also a transition based on when a half-wave reaches about the distance between ears, i.e. 6", again, this is a psychoacoustic issue.
And direct vs. delayed indirect timbre figures very much into disambiguation of front/back listening on the cone of confusion (based on ITD's).
That's just a start. There are bazillion issues, sorry.
But, one smoothed measurement isn't nearly as bad as it's made out to be, assuming it's done right, and properly windowed at different frequencies (no using ONE window, nope). Yes, there is better. Capturing all 4 variables at one point is much better, especially at low frequencies, but also when detecting specular issues), and having measurements about .9 milliseconds apart in space even better than that.
But there's no way I can convey all this, especially here, with some people denying basic psychoacoustics. All we'll get is a rabid flame war, and more confusion.