IMO you're focusing too much on the on-axis response. The on- and off-axis responses, taken as a whole, are much more important to perceived tonal balance than deviations in axial response of 0.5 or 1dB.
Is this true in all situations? (Anechoic, untreated room, treated room)
If it's true in an anechoic chamber and a treated room, and Genelec is flattening out the off axis response at the small cost of on axis response, which makes us perceive a speaker as more accurate, I rest my case: Genelec is no longer blindly pursuing louder, less accurate, less extended speakers over the last decade in my mind. I wonder if any other objectivists have looked at the Genelec "spec decline" over the years and came to a similar (hopefully incorrect) conclusion.