Implicit Riemannian Optimism with Applications to Min-Max Problems