Of course it matters, What I am saying is that; here is a program written in language A, which uses all the tricks in the book to solve this particular problem, here is a another one written in language B, it uses a completely different technique / algorithm and isn't optimized to death. How is it valid to make a decision about performance if implementations and restrictions are wildly different? So to me the site offers very little of value and I can not take any comparison from there seriously.

And as I explained on a different comment, performance of many different language / compilers for this particular benchmark are almost same "If" they use the same technique ( C, C++, Java, Swift, Javascript and Dart.)

But the argument here started as "Hey look language A is faster than language B" , while pointing out completely different implementations, which is, absurd.

> … while pointing out completely different implementations…

danielscrubs referenced measurements which showed similar performance for Dart and NodeJS programs, but did not mention that similar performance.