but is there any science that backs up the claim that everyone’s hearing perception is fundamentally different and therefore an asset for one is another’s liability
This is the kind of question that seems complicated and fraught with paradoxes but is actually simple. The answer like others have said is no, but to address why takes a little more explanation.
Human perception evolved to help us understand our environment. The goals for hearing are to distinguish direction, disambiguate one sound vs. another and pick out important sounds in difficult situations, and furthermore to do all of that both automatically and at will (hearing is an active, not a passive sense). Those are the biological fundamentals consistent between people.
Differences in perception between people in psychovisual studies are termed observer metamerism. I bring it up here because these differences are measurable and account for minor variations in color perception in the general population or severe variations in those that are, for example, significantly color blind or who have macular degeneration. Hearing follows the same pattern, with minor, measured variations in perceived subjective loudness and sharper variations for those with certain types of conditions, like advanced presbycusis or Meniere's disease.
The general problem this concerns is that of personalization.
We know because of the way that headphones couple to ears that personalization is necessary for a subjective sense of realism with binaurally-rendered spatial audio formats, where the goal is physically accurate reproduction of sound pressure at the ears. Recent work by Sean Olive suggests that for stereo content (i.e., a lower standard of accuracy) personalization can be effectively limited to bass/treble controls when starting with a good baseline target.
For speakers, spatial audio (immersive audio is the commercial term) like ambisonics or wavefield synthesis aims at physically accurate reproduction of a defined soundfield. The requirements for speakers and rooms are flat response and anechoic or near-anechoic conditions because the system is meant to provide
all auditory information, including reverberance. As such the speaker has a more background role: the job is not to have any sort of sound at all and to be a pure reproducer. It's telling that the speakers I've seen most often used in ambisonic settings are Genelec coaxials, which have more or less constant directivity across the listening window and a short integration distance.
For normal content relying on phantom centers, like 2-channel, 5.1, etc., the speaker and room have a more primary role in perception. However, while the same general principles apply for speaker design, the relative importance of off-axis sound increases because it interacts with room boundaries. Toole and Olive's work, building on that of others, has helped establish this.
So let's summarize: for very well-designed speakers, when listening to typical formats in nonanechoic settings, differences in individual perception, room interaction and speaker radiation are handled by using simple EQ tilts. If we are being serious we would acknowledge that any speakers falling into this category have minor rather than major differences between them (all of which are measurable) and can be used both for pleasure and critical listening.
Major, personal perceptual difficulties/differences require individual medical intervention. Hearing aids and cochlear implants, for example. Not different speakers.
Edit: Typos.
