Which shows... what, exactly? That a non-engineering approach to music discovery is better? No, you got lucky. That your music discovery procedure was adequate and not worth investing time to optimize? Sure, but that's an engineering judgement.