Ureases and metallo-beta-lactamases are amide hydrolases closely related in function and structure. However, one major difference between them is that the former uses two nickel cations, and the latter uses two zinc cations to do similar catalytic jobs. What is the reason for this choice that Nature made for the catalytic metals? Is it dictated by electronic or structural reasons in the two catalyzed reactions, or some other evolutionary factors? Are both enzymes "perfect" catalysts, as far as just catalysis is concerned, and if they are, then why? Here, we address these questions through a joint quantum mechanical/molecular mechanical dynamics approach and ab initio mechanistic investigation. Five enzyme/substrate systems are considered: urease/urea, CcrA beta-lactamase/beta-lactam antibiotic model, urease/beta-lactam antibiotic model, CcrA beta-lactamase/urea, and di-Ni-substituted CcrA beta-lactamase/beta-lactam antibiotic model. The mechanisms and rates of the metal-facilitated nucleophilic attack are assessed. Both urease and Ni-substituted beta-lactamase catalyze the attack on the beta-lactam ring with the efficiency surpassing that of natural di-Zn beta-lactamase, whereas beta-lactamase is unable to hydrolyze urea. These results suggest that in beta-lactamases the use of zinc does not provide maximal possible efficiency of the enzyme. Thus, beta-lactamases operate by the principle of "good enough"; i.e., the choice for Zn in them leads to a performance that is just satisfactory for its biological purpose but can be evolutionarily improved via replacement of Zn with Ni.