@j_bertolotti in Nature:

"In any discussion about if and how to share data, and how much to share, a large fraction of the debate boils down to people coming from different disciplines, with different approaches to the everyday working of scientific research, and thus different problems that need to be solved."

https://www.nature.com/articles/s41567-019-0625-2

The subjacent issue is that there is no defined scientific method for all fields, disciplines, and subdisciplines (and no need to). So, it is not a good idea to impose the same rules for all.