why is reproduceability the thing?

shouldnt the paper be the math of the argument? the reproduction is reading the following the proof