Vision-and-Language Navigation Today and Tomorrow: A Survey in the Era of Foundation Models