The American Meteorological Society has a good recent blog post on datasets. The focus starts out on making sure every dataset used in any use is properly referenced with the proper citations.
It is important to point out the 2013 AMS Statement on Full and Open Exchange of Data in the meteorology world. While this statement is directed to one discipline, meteorology, similar attempts have been attempted at other disciplines. In the Geomatics/LiDAR world, there are current online discussions going on at rapidlasso, makers of the LAStools LiDAR processing suite, on whether the open *.LAZ format should be supplanted by a closed zLAS format by a market leader and if this serves the long term goals of the technology, and another online debate about having LiDAR data open to the public or kept in a restricted state.
This will continue to be a ongoing discussion in the scientific and technological communities, on what data should be open and what needs to be held back due to other reasons. You can see other discussions on this blog by looking at past posts on the ongoing discussion about data ownership in agriculture areas.