Luciano, thank you for such an insightful article. I’ve searched on Google and I think you are the only person that has related the LLN and User testing. That’s two of us now.

I really think the methodological issue is much simpler, as stated. And many years of UX practice seem to confirm my intuition: Doesn’t matter how bad you setup or run your tests, or how few participants you use, most probably you will end up spotting the key issues of a design.

Moreover I teach design students and even entrepreneurs to run user tests. And the same conclusion stands. I don’t really think these two laws apply in this particular domain.

Your conclusion, however, apply completely in other activities we perform such as metrics and log analysis.