Nothing in nature can ever be described with 100% accuracy by any model. But that doesn't mean models are useless. So imagine why we would use the binary sex model instead of three or a spectrum or what have you.
Simple models are useful, but they shouldn't determine who is allowed to live a normal, productive, life without some very compelling justification. Like the "binary sex model" is handy, but nothing about it makes it obvious that we should definitely and always lock gender (another non-binary model often simplified usefully into a binary) directly to biological sex.
There are only two elements in the universe: hydrogen and helium. The binary element model is 98% accurate.
Is a bimodal distribution, or a somewhat reductive “typical male, typical female, intersex” model, so difficult to understand that we can’t use it? I don’t think people are stupid.
Random but nature related: some birds have four sexes
You may be thinking of species like the white-throated sparrow. These have two morphs with distinct behaviours which lead to there being four mating combinations. Still two sexes though.