Is there any possible to extend the currnet infinite-dimensioal Morse theory, which can be applied to the nonlinear elliptic equations? Almost every paper I read reflects that this version of the theory is (too) well-developed. Is there still any hope to find something new?