I have done research on Evolutionary Algorithm and numerical optimization. It was nigh impossible to reproduce poorly described algorithms from state of the art research at the time and researchers would very often not bother to reply to inquiries for their code. Even if you did get the code it would be some arcane C only compatible with a GCC from 1996.
Code belongs with the paper. Otherwise we can just continue to make up numbers and pretend we found something significant.
In 2006 or 2008 a university in England published some fluff about genetic/evolutionary algorithms that were evolving circuits on an fpga, specifically the published stuff regarded an fpga without a clock was able to differentiate between two tones.
I've spent the intervening years trying to find a way to implement this myself, going as far as to buy things like the ice40 fpga because the bitstreams are supposedly unlocked; this is a pre-req for modifying the actual gate/logic on the chip.
I've emailed the professor listed as the headliner in the articles published about it to no avail.
Nearly my entire adult life has been spent reading some interesting article, chasing down the paper, finding out if any code was published, and seeing if I could run the code myself.
It wasn't until machine learning with pytorch became mainstream that I started having luck replicating results. Just some more data points for this discussion.
Code belongs with the paper. Otherwise we can just continue to make up numbers and pretend we found something significant.