This post follows Peer review is not independent validation, where it is argued that independent validation is one of the pillars of the scientific method. Peer review is only a part of the editorial process. Of course that peer review is better than nothing, but it is only a social form of validation, much less rigorous than what the scientific method asks.
If the author follows the path of Open science, then the reader has the means to perform an independent validation. This is great news, here is why.
It is much easier to do Open science than to change the legacy publishing system.
Many interesting alternatives to the legacy publishing have been proposed already. There is green OA, there is gold OA (gold is for $), there is arXiv.org. There are many other versions, but the main problem is that research articles are not considered really serious unless they are peer reviewed. Legacy publishing provides this, it is actually the only service they provide. People are used to review for established journals and any alternative publishing system has to be able to compete with that.
So, if you want to make an OA platform, it’s not serious unless you find a way to make other people to peer review the articles. This is hard!
People are slowly understanding that peer review is not what we should aim for. We are so used with the idea that peer review is that great thing which is part of the scientific method. It is not! Independent validation is the thing, peer review is an old, unscientific way (very useful, but not useful enough to allow research finding to pass the validation filter).
The alternative, which is Open science, is that the authors of research findings make open all the data, procedures, programs, etc, everything they have. In this way, any other group of researchers, anybody else willing to try can validate those research findings.
The comparison is striking. The reviewers of the legacy publishing system don’t have magical powers, they just read the article, they browse the data provided by the very limited article format and they make an opinion about the credibility of the research findings. In the legacy system, the reviewer does not have the means to validate the article.
In conclusion, it is much simpler to do Open science than to invent a way to convince people to review your legacy articles. It is enough to make open your data, your programs, etc. It is something that you, the author can do.
You don’t have to wait for the others to do a review for you. Release your data, that’s all.
chorasimilarity 
Releasing data is essential, but it is not enough.
I work in mathematical physics, where there are no data, and in most cases no computer code either. In principle, all necessary information is in the article, and anyone can check the results. Nevertheless, it is often hard and time-consuming to do so, even if the article is clearly written (and most articles are not clearly written).
So, in addition to having all the data in the open, we need to make sure that each replication attempt can be shared with the public, in order to avoid unnecessary multiplication of effort. It is not just the authors who must make everything public, it is also the readers – starting with the journals’ reviewers.
I agree, partially. I believe that there should be no obligation. The rest is evolution.