On Benchmarking MiniZinc and the LinkedIn Queens Problem