My intuition was the same, but they are doing this the way you should: Pre-registered, sufficient N for small effect sizes, comparative across countries. I'm not so sure it does not replicate.