The Y axis scale is half that - the 3k dip measured 3.5 dB dip in absolute terms, but the tube traps lowered it 2 dB more.
I'm confused: In your blog you said "The vertical scale (each horizontal line) represents 2 dB." but on the chart you show marked them as 1dB and above you imply that the lines are at 1dB. Can you verify which value is correct, please?
These are Magnepan 3.6/R speakers, dipoles very sensitive to room placement. 1 inch did make a difference, but that "red dashed" line it refers to isn't in the graph because it wasn't different enough to be worth reporting and complicating the graph further.
You answered many of my questions why I am seeing such odd behaviour. You are using a dipole system and such wild anomalies are what to expect from a dipole in a habitable room. As you said even an inch movement alters the response and as wild as yours.
However, if I may analyse your before (blue dotted) and after (blue solid line) this is what I see.
1- You have successfully eliminated two prominent standing waves at 30Hz and 70Hz.
2- However, you have not been able to tame the standing wave at 450Hz. Your treatment have simply moved its frequency up to 600Hz and lower Q. I assume due to the big masses added to the room.
3- The dip at 2500Hz is also not tamed but like above its Q is lowered, most likely for the same reason above.
4- I am ignoring the anomalies between 200-400Hz, and above 2000Hx because they seem to be caused by the dipole speaker's position rather than standing waves.
5- However, I would pay attention to the peak at 250Hz as it nay well be caused by a standing wave.
If I been presented with the blue dotted curve and asked what do I suggest in terms of acoustical treatment here is what I would suggest.
1- Employ two Helmholtz resonator tuned to 30Hz and 68Hz.
2- Place an absorber that has coefficient of at 0.8 at the 300Hz-1000Hz range.
3- Once the above in place test to see if there are any more important standing waves, especially the one at 250Hz. If found and if they are higher than 100Hz, employ a single resistive (absorber) tube trap.
Once again, I congratulate you for the work you have done and the fact that you thought of treating a room before using an EQ. Bravo!