The RGB you'd typically use when designing or developing something does not really map very well to human biology because it's just an approximation that happens to work well with the technology we currently have.
For example, the dynamic range you can express with standard 24-bit hex RGB colors (e.g. #ff0000) is way smaller than what the human eye is capable of perceiving.
Also, using mathematical models to create palettes typically relies on a lot of assumptions. For example, #ff0000 is not perceived as equally bright as #00ff00 by a human being, yet looking at it represented like this you'd assume this to be the case.
I think it is possible to build a mathematical model that creates pleasing color palettes, it's just that they're often built on approximations that aren't really accurate.
For example, the dynamic range you can express with standard 24-bit hex RGB colors (e.g. #ff0000) is way smaller than what the human eye is capable of perceiving.
Also, using mathematical models to create palettes typically relies on a lot of assumptions. For example, #ff0000 is not perceived as equally bright as #00ff00 by a human being, yet looking at it represented like this you'd assume this to be the case.
I think it is possible to build a mathematical model that creates pleasing color palettes, it's just that they're often built on approximations that aren't really accurate.