Real World Diagnostic Benchmarking for Modern Many-Objective Evolutionary Algorithms