That's fair, if you've been benchmarking but haven't made the benchmarks public / repeatable yet. Too used to software where the authors claim it's fast with no proof or based on heuristics like what language it's written in :-)