That's the kind of thing that real bass traps should help with. Kill (or reduce) the reflected waves and you kill (or reduce) the standing wave nulls/cancelations (and the anti-nodes/bumps).
...You can fix the bumps with EQ but it takes "infinitely large" subwoofers and "infinitely large" amplifiers to overcome cancelation.
I don't know what's going on in the 50-90Hz range. The rooms 2nd length mode is at ~82Hz, but it doesn't seem responsible when looking at the various charts and filtered IR, because it looks "normal", just quiet. But also the 1st width and ceiling mode are in the same range (~60Hz and ~70Hz), so they might be contributing?
Here's some results of my right speaker (the volume of REW/system was the same, but I don't know if the gain was exactly the same in EQ APO, but <38Hz looks the same, so probably the volume was the same);
There's obvious improvements across the board there. My room was totally untreated before with brick walls and floating timber floor, so not hard for there to be improvement. Is this improvement better or worse than normal panels, I don't know? Do normal acoustic panels achieve such effects at 100Hz? I've not looked closely but I don't think they do.
Here's the SPL, with sub included (no EQ or LP on it);
The sub looks great below 60Hz, but it too is suffering from 80Hz problem, and considering it is roughly in line with the speakers, is suggesting room mode. Maybe it makes sense, because the back wall is largely untreated, so there's a full force reflection coming and mixing with a diminished wave that came off the front wall?
Also the sub has issues above 100Hz lol. Good thing it doesn't need to operate there. The ~116Hz peak is the 2nd width mode, which makes sense because the room is basically not treated in that direction. I haven't thought about the big hole at ~180Hz.
There's 2 other interesting changes in that data;
1) the massive rise in clarity at ~200Hz, which also has a corresponding dip on RT60 graphs.
2) the lack of SPL in 400-600Hz range that corresponds with a blip on the T20 graph, and possibly it is dragging down the clarity graph too.
There's much to ponder here. Looking at filtered impulses of around 500Hz, it seems there are a lot of new reflections, which is hard to fathom (maybe the shelves?). I did see a post or 2 from other forums about mixing in some heavier density to target different frequencies, which I might consider. I wonder if just jamming in another piece of 2 in the very middle will help, as a pseudo "acoustic panel"? I might experiment with this.
Also on the experimentation front, I pulled down the 2 panels I put up near the door. These would have helped when my room was untreated, b ut now the state of things is completely different. So here I measured with 2 panels, just the bottom panel, and then no panels. It seems those panels were making things worse at ~500Hz. Very interesting. I will have to try them in other places and see what happens.