Can you post a link of your source for this comment? The depth of the music has nothing to do with the shape of the diaphram. Regardless of the type of drivers usef the basic principle is the same. Sound is produced by the vibration of the driver's diaphram.
Depth is just a perception how a driver produces specific sounds relative to each instruments in the scene. Say a cymbal sounds softer than a guitar or vocals create an illusion the drums is behind the guitar. Now it depends on how soft the cymbals relative to the guitar creates the virtual depth. Now if you factor in the left and right channel balance give you the coordinates of an instrument on a virtual soundstage.
I have never read from anywhere the shape of the diaphram influences any aspect of a drivers ability to produce sound.
That is my contention based on reasoning and based on different presentation of the soundspace of headphones of different technologies that I heard. What I heard when listening to planar magnetic technology headphones was a flat two dimensional arrangement of instruments. Planar magnetics have good left to right dimension of soundstage and the height of it but the depth is lacking. And this projection in depth has nothing to do with whether an instrument's sound is projected as soft or as hard. My view is that a gradation in softness of sound, or the lack of it, that results in perceived distance in depth, or the lack of it, cannot be an answer. Or possibly can be one factor, but then it would mean that the planar magnetics cannot project the sound with a variation in 'softness' as you say. If that is so then we are back to square one, the planar magnetics have uniformly inferior projection of three dimensional sound compared to the electrodynamics and there has to be a reason why that is so. The concave surface of a diaphragm of the electromagnetic headphones does not vibrate uniformly across its width, the tension of the membrane close to the inside ring differs from the tension close to the outside and that might provide an answer. Another factor might be that there is also a micro-time delay of the sound waves coming from the inside of the concave shape membrane compared to the waves coming from the outer edge. Somehow either of these two [three ?] phenomena or both of them together cause good projection of 3D soundspace of the electromagnetic drivers/headphones. Compared to an electrodynamic driver the diaphragm of planar magnetic headphones has a flat two dimensional shape, has a uniform tension across its width and height and vibrates evenly. The fact remains that the planar magnetic headphones do not have good three dimensional soundstage. The shape of the diaphragm has to be A reason for this problem, and maybe THE reason. I have not heard electrostatic technology Stax headphones but I read some comments that regardless how good Stax headphones are their projection of soundspace is not great. Stax headphones also have flat two dimensional diaphragm.