I'm just going to say this. It's a great sentiment, but unless you're in the thickets you just don't realize how hard it is. I'm trying, trying to maintain an open lab notebook with my nonprofit. It's hard to remember to add details in edgewise. Sometimes you just want to do the experiment, and you're so damn tired that when it's done, it's hard to sit down and write things down. Sometimes, the lack of tooling to do this makes it harder (ever tried to export an HPLC trace and then convert it to a format so that you can view it on a web page?).
Something like benchling - is a positive step towards better. Zero friction (which benchling doesn't have yet) reporting is an absolute must. Run a PCR? Ok, that run gets uploaded to the notebook and filed in the correct experiment. Run a gel? Ok, the gel image gets uploaded to the notebook and filed under the correct PCR run that it corresponds to... Etc etc etc.
Even in places like theoretical physics that aren't connected well to experiments, the literature suffers from the lack of copyleft. Writing good review articles and textbooks are huge, collaborative undertakings that are crippled by the impossibility of directly building on the text of others. Especially in fields that use the arXiv, open access is important not so much because it makes the raw information accessible -- it already is -- but because the journals won't care about the further jump to copyleft. Open access and copyleft are financially equivalent to them.
Science magazine is in a great position to push for changes. They could force their contributors to publish all results, including failed trials, open lab notebooks, etc.
I think the funding agencies would be a much better place, since they have a near-monopoly in many fields. And they decide "how much work is enough" for the money they hand out.
Something like benchling - is a positive step towards better. Zero friction (which benchling doesn't have yet) reporting is an absolute must. Run a PCR? Ok, that run gets uploaded to the notebook and filed in the correct experiment. Run a gel? Ok, the gel image gets uploaded to the notebook and filed under the correct PCR run that it corresponds to... Etc etc etc.