Wow...if that frequency response is accurate, we're worried about how to treat a back wall, that even if it is totally reflective would create an ISD out so ~37 ms?
You need to figure out the nulls and treat them.
But first, WTF is it with the 3 stereo pairs of speakers on the console bridge???? And why are what appear to be identical vertically mounted speakers set on an incorrectly horizontally mounted speaker positioned SO high where the tweeter almost couples with the ceiling????
You aren't running both vertically mounted pair simultaneously are you???
So before you do anything else, you may as well actively resolve a few problems and anticipate a few solutions.
You have the space, so pull the console out from the wall and PROPERLY mount a single pair of the speakers on stands located ~3 foot behind the console to put them on axis with the listening position and to minimize work surface reflections!
AFTER you do the, generate a waterfall plot from 0-300 Hz and a left/right ETC (or better yet, first mark the position of the center of your head in the listening position with a small string and washer acting as a plumb tacked to the ceiling. This is where you will place the capsule of your omni mic for ALL measurements, now or in the future. Then just generate the frequency response and impulse responses and post the 'raw' source files - allowing us to import and massage them in REW for various views.)
And from the various test results identify the source of your nulls and treat them appropriately. But first things first.
Also, noting you have what looks like one LENRD on the floor in each front corner and only very small corner traps in the front and nothing in the back corners, before you worry about a back wall, you are going to want to address room modes most likely with substantially beefed up corner traps such as superchunks with 2' wall exposure and an ~34 inch wide front face.
But one can tell much more with the test results.