As I understand Dolby's use of their "Atmos" terminology it is not a product or a discrete technology. It is hard to pin down the specifics other than its reliance on "meta-data". To my mind that meta data is sorta like the EXIF data that comes along with digital photos. The "Atmos" system then uses the meta-data to map the various channels as laid down by the studio to the various outputs that are available in the cinema. Each film would have the director / sound designers meta-data for each "element" of whatever was recorded. This would allow for far more specific placement in both "real space" and the "virtual sound field".

This is VERY different than Audyssey which is way a measuring room response and making corrections via filters to attempt to make the sound reproduction closer to ideal. Trinnov similarly measures room response and adds in the ability "focus" the reproduction in a way the maximizes the apparent separation / symmetry even of "compromised" speaker placement...

The "future home product" aspect of Atmos certainly seems theoretically viable and if it leverages the sort "refocus" aspects of Trinnov the ability to create "virtual" source locations that could simulate whatever the filmmakers intended. It ought to be possible, just like with Audyssey's various "higher resolution" variants, to be better able to do the "digital mapping" with greater precision but it is not clear yet whether the "raw metadata" is the limiting factor or whether some "processor" of some specific CPU rate / memory bandwidth would serve as the limiting factor...