Trinnov is the only [AFAIK] processor capable of rendering 24.1.10 Atmos content + 128 objects.
Irrelevant. Fixed objects and bed channels do not render to speakers like Front Wides despite correlated information present between the front mains and side surrounds in most movies and surround recordings. $20k-48k and Atmos still doesn't use those extra speakers properly.
This purely Dolby's fault. Scatmos gets around the problem as does Neural X from base 2.0/5.1/7.1 soundtracks. DTS:X Pro doesn't exhibit the problem either. It's directly related to Dolby's implementation of object rendering (particularly with fixed objects which unfortunately are used a LOT in movie soundtracks) versus bed channels.
Re Matrix - it is in Expert Menu and you can basically route whatever you wish to wherever you wish, sum up channels, multiply them, delay, gain etc
Routing and summing do not necessarily mean Matrix decoding or logic steering. A mixer sums, but it doesn't give discrete output for a summed center.
No real reason to do it, as you have couple of other options - e.g. 3D remapping or arraying of speakers and of course Auro3D
None of those help the above described Atmos issue with fixed and bed objects not using speakers like Wides WHATSOEVER.
Re Logic 7 - this is not an issue to re-create in Trinnov, e.g. for 2Ch music and with 9.x.7 setup it takes Logic7 couple of notches higher, But you also get very similar results with adjusted Auromatic upmixer, you just need to have proper speaker layout
Um, no. Auromatic isn't in the same galaxy as Logic 7. Auromatic is nothing more than a cheap reverb copy system. On the plus side, it doesn't screw with the front sound stage, but it does little more than a cheap DSP reverb effect. It's not even close to what Logic 7 does.
Neural X works well with 5.1 and 7.1 sources, but so well with stereo. So the idea would be to use the Trinnov 7.1 inputs with an analog Logic 7.1 output and run Neural X on it to separate out an overhead layer from the 7.1 base. That is something I'd love to play with.
Trinnov anyway recommends heights, not Atmos tops.
Trinnov supports all 10 overheads simultaneously and more is better for precision imaging across multiple seating rows or a long room. Six is usually sufficient for most medium to semi-large rooms.
I use reverse polarity Scatmos for my overheads. Instead of steering discrete output to the center output, reversing the polarity sends it to two or more surround outputs instead (you then right the polarity on the speakers). This gives me Tops from Heights for 8 overheads (and quad mono VOG in Auro-3D). You could do 10 overheads with a 6.1 PLIIx AVR and my room would normally be small enough to get by with only six overhead speakers, but my room has a steel beam box where Top Middle would be mounted so Heights+Tops gets around the problem very well.
I have heights above ear layer speakers angled at MLP and speakers are capable to go down to 60Hz without compression. I run heights bit hotter to enhance immersion and you need to carefully adjsut bass management in order not to get too pronounced bass, once you get into stronger Auromatic setting.
Naural X - don;y use st all, unlistenable, same for for Dolby surround
You're the only person I've EVER heard from thousands that has claimed Neural X is unlistenable. I can only surmise that there's either something wrong with your Trinnov or the way you hear.
IMO, arrays, summing etc is no substitute for discrete steering to more speakers. Neural X can make a 5.1 version of an Atmos soundtrack (e.g. The Meg) sound almost exactly like the Atmos version. Dolby's own Pro Logic II steering can make a pretty discrete center between two analog inputs (no leakage with fully correlated in-phase inputs).
This is the same basic thing the panning renderer in Atmos does when adding additional speakers between say the Front and Rear Heights (Top Middle Left & Right). It renders based on the object position rather than the amount of correlated panned content, but panning is panning.