print more statistics when parameter-optimizing

during evaluation

• number of benchmarks: solved, failed, todo, remaining allowed failueres

when presenting a population

• more info on singular benchmarks (solved by few)
• and on singular configs (solving most, or most singulars?)