A good room correction system should allow users to tweak the target curve to their preference. However, first impressions matter, so it is helpful to start off with a default
target curve that sounds good right out of the box, even before user adjustment. To that end, several room correction designers (Harman, Dirac, Lyngdorf) have done research into finding a good default curve. Harman, for example, compared several room correction products and found that perceived flat response was preferred over measured flat response. Likewise, Dirac has a default target curve that tilts 5dB down from 20Hz to 20kHz.
Lyngdorf wanted a default curve that sounded natural to listeners and looked into what "natural" might mean, documenting the research in two papers (see previous post). The papers concluded with an explanation of why it is important to measure around the room to get additional frequency response information:
"The additional information is also used to automatically calculate a target function, which suits the loudspeaker actually used, which opens up the possibility of a fully automatic room correction system, where no user interaction is needed. This system also provides a more natural timbre to room correction systems by recognizing the fact that part of the influence of a room is perceived to be natural and should not be removed by a room correction system."
Based on the above, "natural" turned out to be different for each loudspeaker and room. Removing some of the room's influences, like the naturally occurring hump in the low frequencies, sounded unnatural to listeners. Which meant that a natural sounding target curve was a matter of tracing a line through the measured frequency response of YOUR speaker in YOUR room and conforming the peaks & dips to this line:
Your speakers still end up sounding like your speakers, since the default target curve was based on their frequency response, but the room's unwanted
contributions (peaks & dips) are removed (as much as possible). Now, this doesn't make it the best room correction in all of Christendom, but it does explain why the results are often described as natural.