That worked with creative sound-cards for ages. Nothing really new. On a stereo headset you can here a sneaking player behind you (since battlefield 2).
And again, this only applies for stereo headsets. This does not work very good with 2.1, 5.1, 7.1, ...
Yes, you can make more calculations now, but like e.g. Dolby expressed, more sources is not better. At a certain point, it just creates more noise.
Audio processing is not like graphics processing where you can scale almost without limits. In the end, there is really not much more to gain. Only under very specific circumstances.
Even MS tried with kinect 2 scanning your room to make your audio entertainment better (and also used the mic in kinect to test the settings). In the end, it didn't do much. Audio output is really heavily depended on the audio setup and the person which uses it. Adding much more processing time into this is more or less a waste of time and in the end, game developers will use the cpu (like in all generations before even if there was audio hardware).