I read the blog post, but it just seems to provide a single example of a case where some particular piece of Objective C code was faster than the equivalent pie

I read the blog post, but it just seems to provide a single example of a case where some particular piece of Objective C code was faster than the equivalent piece of code in Swift. Maybe there's something I'm missing here... what should I be looking for in these blog posts?