Wouldn’t the proper solution to this be to fix how data is published? Instead of making the actual paper unreadable to help people replicate it, maybe require a standardized structure to provide the necessary data and code to replicate the calculations?