He wasn't wrong; it was just discovered that it was an approximation accurate for a specific domain rather than universal, once we began to understand relativity. Classical mechanics are still very much 'correct' and useful within the non-relativistic domain.