On the Evaluation of (Meta-)solver Approaches