Would crowdsourcing peer review work? Probably not, at least in fields I know anything about. What you need for a peer review is the guy (the gal). The gal or guy who knows the lay of the land, who can compare the article with existing literature. Crowd sourcing works best with lowest common denominator tasks. If I were to ask someone to read an article or chapter for me, it would not just be any old person. Of course, you can also have a reading by someone who knows nothing about your particular topic. Then the reading serves a different purpose, because you have to be more clear about your assumptions to convince that kind of reader.
The reader slightly out of your field can be good, to bring another perspective. So I might be a good reader for your paper on French poetry, say. I wouldn't have any hobby-horses about the subject matter.
But generally you want the guy.