Dynamic Range in Headphones, or lack thereof

It is a given that a higher sensitivity speaker will have a larger SPL out range for a fixed set of Voltage In values compared to a lower sensitivity one. To me that follows from the definition of sensitivity.

The question, for me, would be if a higher sensitivity speaker will always be more linear than a lower sensitivity one, in the lower sensitivity speaker output SPL linear range. That is, THD+N at a reasonable operating frequency for the two drivers at levels where the two drivers should operate.

My expectation (or bias or whatever) would be in line with @OJneg's: Speakers will show good linearity within reasonable levels.

As far as cross over regions showing compression. I think that's another story.