For some things, yes. For others - not really. And there have been groups, historically, that have not been served well by "Everyone has access to this data". Given the need for their consent and cooperation, I generally side with what my subjects want their data to go toward, rather than some nebulous ideal.
In today's world of github gists and python PyPI's and R's CRAN, there's no excuse to not document the entire process, in addition to raw data.