Well, yes. Another way to look at the benchmark game is "if I'm not using a fast low-level language, theoretically what speed gain might I expect if I wrote a portion of my project in X language and called out to it (ignoring data marshaling)?"

It would be interesting if entries were categorized into idiomatic and non-idiomatic representations of that language, but that would also require people agree on what is idiomatic for that language, which is easier said than done in a lot of cases.