If a model is accurate enough and practical, there's no reason to stop using it.
Good science is pragmatic and not just a pseudo-intelectual masturbation.