Sorry, I'm not trying to be argumentative, but TimVG and I were simply responding to your earlier point that:
"in reflective rooms there is no HF downward slope, it can be straight as an arrow from 1KHz to 20KHz."
I was just trying to demonstrate that the reflective room bit is, to the best of my knowledge, incorrect.
You are definitely correct that a speaker can have a straight in-room response from 1kHz to 20kHz, but that's a function of the speaker's directivity. It being a reflective room isn't going to majorly affect the overall tilt of the response, except for above ~8-10kHz or so.
To perhaps provide more context on this matter, the aforementioned room where I measured the Buchardt S400 is what would be considered "a reflective large room." Where measured, the speaker was about 2m away from the nearest walls, and about 7-8m from the furthest one. Ceilings were about 5m high.
Nonetheless, you can see that the speaker maintains a tilt in its actual measured in-room response. The reflective room did not help the speaker remain flat, despite the fact the S400 I measured was very flat on-axis anechoically.
Here is a graph comparing the on-axis to the PIR and the actual in room response. Smoothed to 1/3 since trends are the only thing that matter here.
View attachment 133428
On the other hand, the D&D 8C has something close to constant directivity throughout most of its frequency response, so instead, it has a flatter, less tilted PIR and actual in-room response.
View attachment 133431
The KEF R3 showed a bit more devialtion from about 8-10kHz, but nonetheless tracks the PIR better than the axial response.
View attachment 133435
So a reflective room will not straighten out the response. These weren't super controlled measurements, but in two of three examples, the actual in-room response is actually a little more tilted than the predicted in-room response above. It is the directivity that will affect the slope of the in-room response above all else between 500Hz ant 10kHz.