Towards leveraging collective performance with the support of MPI 4.0 features in MPC