Hacker News Reader

Someone knowledgeable enough about movies can read a script and know of it's good or not without needing to see it actually produced.

Here, it's possible to read the code and know what the program does sufficient to critique it for what it is.

But he did read the "book" (source code). But ignoring analogies, can you cite a specific benefit to running the benchmark when discussing parsing correctness?

That's not a very good analogy, you can understand code without having to run it.