This is not a complete answer. But the point of systematic open science in the first publishing is, among other things, to prevent scooping, yes? 

Consider that scooping (defined by this sentence implicitly) is possible iff one publication is obscure or informal and another published later without crediting the earlier source they were aware of is much less obscure and claims the result. A systematic if informal system like arXiv prevents scooping actually, as you would imagine. Even MathOverflow posts have been [cited in papers, not scooped][1].

I would have difficulty finding an example of scooping in internet publication because of this reason. It's easily discovered and I don't think that people who participate in open science do it because it undercuts the goodwill foundation of goodwill (an intangible asset that goes on a balance sheet and can be in other circumstances a large part of the value of an enterprise) of open publication and easy mutually beneficial discussion.

Allegations of scooping arose mostly in the eighteenth and nineteenth century when most print publications, even well known ones, were difficult to obtain or check. Likely remarks about such cases are where the fears of being scooped originate.

  [1]: http://meta.mathoverflow.net/questions/617/best-of-mathoverflow