To be honest, I totally ignore that predicted room response graph because it is literally unpredictable parameter changing significantly from case to case.
I used to watch Andrew Jones interviews , who worked in KEF, TAD/Pionee and ELAC. He has rare skill of explaining complex matters in simple way, and as an engineer, despite being commercial guy now, sometimes he spills out some " behind the scenes truth".
Room response was one of the "pains" he had, when was designing the speakers.
He had to consider some single average level.
So I totally ignore it for myself in Amir's measurements, since it has 0 relation to my particular case, unless I build a dedicated listening room and apply some treatment.