I think it depends on what measurements we are discussing. FR is easily the most discussed and influential. Measurements are objective... words like "Detail retrieval" "Note weight" are not, thus, have no real meaning in science, thus, theres no measurements for those aspects.
FR and THD are probably the most important aspects but since headphones are so intimate with the ear, Timbre is likely a huge factor. If one could ever hear the material of the diaphragm or spectral decay character, it would definitely happen with headphones
Sound stage within a headphone is a strange concept... anything apart from the original is an augmented reproduction. M/S processing and crossfeed would manipulate this. In loudspeakers, directivity and time alignment would affect this in combination with room performance . Headphones need no time alignment and there is no directivity to think about since there is no room influence.
For closed back headphones... if using EQ to flatten the FR, one could use a multitude of quality headphones to achieve a high level accuracy.
Since there is only one representation of correct. As you reach the top of performance the products will sound more similar than not...
The trick is discern whether you are interested in true accuracy or preference.
FR and THD are probably the most important aspects but since headphones are so intimate with the ear, Timbre is likely a huge factor. If one could ever hear the material of the diaphragm or spectral decay character, it would definitely happen with headphones
Sound stage within a headphone is a strange concept... anything apart from the original is an augmented reproduction. M/S processing and crossfeed would manipulate this. In loudspeakers, directivity and time alignment would affect this in combination with room performance . Headphones need no time alignment and there is no directivity to think about since there is no room influence.
For closed back headphones... if using EQ to flatten the FR, one could use a multitude of quality headphones to achieve a high level accuracy.
Since there is only one representation of correct. As you reach the top of performance the products will sound more similar than not...
The trick is discern whether you are interested in true accuracy or preference.