Yes, headphone simulation does work. but consumer has to have his head, his auricles and most important his ear canal 'sampled'.
then it does work. but we're talking about some david griesinger in his lexicon labaratory complexity and effort.
It's not done with taking a picture of ones ears and send it to some plugin company.
And then we need processing power. For every object in that atmos mix these renderings have to be done and for the beds as well.
That could drain a smartphone battery rapidly.
And I know, your gaming teenager says it works with his corsair headphone without all that shit, that he can hear enemies from behing.
after gaming for like 8 hours a day making his brain learn those new hrtfs that are not his.
I know I sound like the guy who says the party is over before it even started, but to me it's groundhog day all along.
It will not be mainstream ever. As much as we want it to be.

 

Be in the mainstream or not one day, I am having fun at re-mixing some music composed by another artist on Dolby Atmos using the stems and sounds of the original songs, then I render the downmix to Binaural and share them on Youtube.