Could wireless USB be a case of "if all you have is a hammer, everything looks like a nail"? [1]
I.e. the effort was driven by the USB-IF [2] that happens to be more hardware than software oriented. So they were eager to deliver a solution based around a new chipset that could be adopted immediately by anyone interested.
This failed to account for adoption friction/lag, and the era of ARM-based SBCs and WiFi proliferation which was already dawning (e.g. iPAQ handhelds were available at the time [3]).
So, they ended up with most of their envisioned use-cases [4] being covered either by SBCs, or by Bluetooth. At least in retrospect, standarizing a pure software solution like USB over IP, as an added-value proposition for the USB standard, would have made more sense.
[1] https://en.wikipedia.org/wiki/Law_of_the_instrument#Abraham_...