The risk here is that the original code contains a bug that is inadvertently reproduced by the replicating scientists after reading it. It can easily happen in some fields that some computation looks very plausible in code, but is actually incorrect.