We monitor in surround on location, so typically I take my height elements and route them to a separate quad subgroup on the Pyramix and monitor them on their own, checking for continuity of imaging with the 5_0 mix and over all front to back blend. I then blend them with the mains and check that they are not fighting. Then, for the most part, I periodically check to make sure nothing is wrong and mostly ignore them during tracking.
Immersive is really a post production intensive process. There are so many variables in how you mix immersive. Virtual room shape and size; box, dome or slant top; and how you monitor the material (Studio or stage, 5, 7 or 9 channel lowers, 2 or 4 channel uppers and mono or stereo LFE.). Like was mentioned above, theatrical has pretty much been sorted out, but music is still the wild west. For the time being, i'm using my room and the comparison of theatrical with the theater to get a good sense of the overall imaging options. Beyond that, probably the greatest listenership will be with the Dolby headphone virtualization, so that is part of the process as well.
Atmos is not a high def format, and as such has its own set of limitations. Today, ATMOS production is limited to 48k and for that reason we are treating it as a delivery medium and not a master. I'm doing all my production at DXD or 192 and then at the end creating a 48k mix master project for ATMOS delivery.

As always, YMMV.
All the best,
-mark