If the mixing engineer cares enough, he understands that not everyone is going to have full 3D systems. Proper care also involves making sure the mix will sound great in 2D. Once you do multiple mixes, there are things that can be done to ensure the mix sounds great on multiple systems. Mixer intention doesn't have to be as limited as you make it out to be according to at least one engineer I went back and forth on the subject with.
I would hope a good mixing engineer would test the end results on at least headphones and more than one configuration with music sources, IMO (maybe not realistic to expect for an hourly worker?). When I made my own stereo album, I did my own mastering and I tried it on 3 home systems (including one high-end), my car system and headphones and made adjustments accordingly to sound great on all of them (Pink Floyd fans seemed to think it sounded up to snuff when I was trying to get second opinions). I've personally purchased too many music albums where it only sounds good on a cheap system with the bass cranked all the way up and so I decided to create my own CD Review site back in 1996 while still going to college after getting my Carver AL-III ribbon speakers. It was called "The Audiophile Asylum" (not affiliated with the "Audio Asylum" that oddly showed up within a couple of years of it).
One of the worst offenders in memory was Firehouse's original album that sounded GREAT on my old El Cheapo "all-in-on" system when I was a teenager and pretty good in a stock Chevy, but my god, it was horrible on my Carver AL-III ribbon speakers when I got them and those speakers generally are at least tolerable of mediocre recordings (no better/worse than on anything else, but much clearer/revealing on high quality recordings). I put that in my review of the album, which I used to give two ratings, one for music quality and one for sound quality. I focused on finding good music with great sound quality (for my own collection) and figured I'd review them as I went along with my favorites from my existing catalog, but I did include a few albums I really enjoyed like Firehouse that didn't impress me with the recording quality and I even got an email from someone I'm pretty sure was one of the band members (he was pretty offended, until I explained what sound quality meant in the reviews and it wasn't the music itself and then he seemed much better about it). Between moving to a new house and job (and different Internet service that was hosting the site) and honestly running out of new music I liked as the industry shifted into pushing more Rap/Hip-Hop rather than rock music, I just decided to end it after ten years (2006). I hadn't had many new reviews for a couple of years anyway.
But it just seems like it would be pretty simple in a proper Atmos studio environment to set the playback to use lesser configurations to see how it sounds in those modes (from Stereo to 5.1 to 7.1 to 5.1.2 to 5.1.4 to 7.1.4 or more). Even on my home Atmos system, I can set it to do anything from stereo to 3-channel stereo to 5.1.2 or 5.1.4 all the way up to 11.1.6 including true Auro-3D 9.1. It's just a matter of pressing a few speaker switchbox buttons and changing the amp assignment. It's actually interesting to compare how some movies sound in 5.1.2 or 5.1.4 versus 7.1.4, particularly Auro-3D vs. Atmos versions. Oddly, I found that the speaker layout had way more effect than the mix itself (since the Auro mixes were probably based off an Atmos master). 5.1.4 Atmos sounded virtually identical to 11.1.
Auro-3D recordings played back over the same 5.1.4 speakers and Auro-3D using expanded 11.1.6 (matrixed FW and SS#1 and "scatmos" center-extracted top middle actually work with Auro-3D and give it 9.1.6 playback) sounded pretty close to Atmos renderings over the same speakers (save rear surround which didn't go back quite as far, but from the 1st row sounded pretty similar as the matrixed sides expanded it to 2/3 the way back into the room and most movies do very little with discrete rear surrounds for some odd reason and Atmos hasn't changed that a whole lot, with some exceptions in some scenes, of course. I'm not sure why that is because rear surrounds placed further back can do some amazing things in Atmos in layering sounds behind and even through you while still being able to put sounds in the back of the room (unlike upmixed 5.1 which puts all in-phase behind you sounds in the back, with nothing right behind your head or whatever).
There's even an
Atmos demo called "
Silent" that has a sound of a creaking hand-crank (that sounds kind of like a squeaky cart on wheels) being panned back from that starts behind you and passes right through you if you're sitting in the middle and your system is set up right and goes to the front center of the room; it's pretty freaky. 5.1 played back through DSU or Neural X would cause it to "go around" you to the sides instead of straight down the center middle of the room at ear level). Unfortunately, most real movies don't move objects through the middle of the audience at "ear level" as most theaters would probably make it sound overhead anyway as they tend to put the side surrounds 2/3 up the wall) and so they're more likely to choose the ceiling speakers for straight middle front-to-back sounds. There are some exceptions for placement, though. Jurassic Park in DTS:X puts the raptor in the cage at the beginning pretty much right behind your head in 7.1.4 or greater if you're in the front of the room with the sides just behind you with other sounds going to the rear center instead. The original 5.1 DTS soundtrack upmixed had it in the back of the room (which for 5.1 systems would likely be in the same spot, but there was a clear difference here).