On the analysis of the (1+1) evolutionary algorithm for the maximum leaf spanning tree problem