No matter what you use to create the sound you are only sensing what happens at your two ears - so if headphones could faithfully recreate what you would hear from any number of speakers you could recreate the experience. HRTF and other computational methods try to do that with various levels of success