Enhancing interpolation and approximation error estimates using a novel Taylor-like formula