It's not the risks they take (which was your question), but it's an example of some value that they add to the process which usually gets overlooked.
And academia is already about networking; collaborations, at least in CS and Mathematics, are a big part of research, so finding the right set of reviewers shouldn't be too difficult for editors of a journal, given that these editors are probably well-placed people in the field.
It's both an insult to what publishers do bring, and something of an own-goal: if you want to replace a scummy operator, you have to understand what they bring to the table so that you can replace it with something better. It's identical to the people who scream about Apple and Google taking a 30% cut of app store sale - these people are clearly clueless about business and are shouting to the world their ignorance about what that 30% gets you.
> And academia is already about networking; collaborations, at least in CS and Mathematics, are a big part of research
Academia also has a lot of political backstabbing and competition for dwindling grant money. But Elsevier's role isn't finding collaborators - it's finding reviewers, which is a distinctly different thing. You also can't find blinded reviewers yourself or via known associates, because that takes away the blind.
Reviewers themselves get paid nothing at all.
In CS, the journal publisher/ conference proceedings publisher does nothing to organize the editorial board or pick reviewers.
For conferences, there are a couple of Program Committee chairs, and they pick the PC that then review papers and solicit external reviewers for help. Once again the publishers have no role to play.
Electronic journals such as Discrete Analysis in mathematics are already doing a decent job of offering cheap open access to articles, and have a solid board of reviewers.
This is also not true. It's usually journal editors who find reviewers. This is mostly unpaid work as well (publishers tend to chip in a bit for editorial board meetings, but that's about it).
Scientific publishers provide very, very little value to the scientific community. The reason why researchers want to publish in Journal X is that it has a good reputation, which is mostly a function of the editorial board's quality standards, and even more so, of Journal X's past publications (often dating back to way before Elsevier or whoever else actually acquired it).