What could external processors really do better than just emulating a speaker room. Sure, SBX does it best so far with best cohorence, least quality loss and least artificial reverb but in the end it's still a speaker room and it can'd do what you wish them to do: like elevation. We could add more speakers like 11.1 or 22.2 (afaik there's support on the Xbox One at least) and approximate even better and but I doubt anyone care's enough yet and I also doubt that such a solution would be affordable below the price of a Headzone.
Read about AMD True Audio, GenAudio (in combination with True Audio, FMOD, WWISE) and Sony's GDC conference yesterday about VR and binaural audio. Those approaches attack from the right angle. True Audio only just begun to start, we can't expect a lot with only one title. More to follow, more experience gathered by devs etc..it'll take time but it was important that it appeared.
Proper HRTF or binaural audio relies upon geometry, coordinates, material informationen etc..directly from the game engine mixed to the stereo output of your choice via (FIR) filters, algorithms and and not through speaker room HRTF like Dolby Headphone and Co. Sure we can fake elevation with prerecorded samples but the illusion blows away once you change the direction of your view or turn your head during VR. This is the only way to have a believable soundfield.
So far there is no PS4 game whose "headphones" option outputs such a binaural mix. VR games for the PS4 will do it, the question is if we get the mix only from the VR or also without it through optical out or HDMI. If not then we better tweet, mail the hell out of them.
You still have to rely on external speaker room processors so far.