It has been the main argument against unifying scholarly literature and data in a single database that monoplies are always bad. I won't go into any political arguments here. Markets have their place, but they also have their limitations. In a recent email exchange with a fellow scientist, we fired these arguments (among others) at each other.
He wrote:
I replied:
If competition were really so important, why does each university have one library system? Why hasn't that been outsourced to whatever provider provides the best library service? What is there to corrupt when all that is required for publication is scientific soundness? What is there to be corrupted when where you publish is irrelevant, but what you publish? And besides, with currently 1.5 million papers being published every year, you need so many people that if for some reason you get into a feud with someone, there's plenty of others to chose from. And if someone publishes nonsense, post-publication review takes care of that.

So I don't buy into the competition argument at all. Markets work. But nor everywhere and not for all purposes.
To which he replied:
But nobody owns the standard. The system depends critically on a standard to which all libraries subscribe. Think of it like email. Who owns smtp, or http? There is one single system, yet, no mafia, only spam (and that's what peer-review is for: not to eliminate, but at least to reduce spam).
Happens now and will be less later - because just publishing isn't rewarded. What is published is being rewarded, not just publishing.
A standard in publishing is as much like a monopoly as http or smtp...
Of course existing publishers can join the standard, but why would they? There's no financial incentive for them anymore, because libraries provide the same service for less.
Yet people love to publish papers falsifying someone else's papers. Science is a single, gigantic post-publication tool, that's how it works. What else is a paper other than a comment to all the papers listed in the references? It's just the formats that vary and that just takes acclimatization. With a common standard and a single database, all we're doing is just tapping the potential of something that has been there all along: multi-level communication. What I'm advertising is not the creation of something new - it's merely trying to stop historical baggage to prevent the potential discourse that is inherent to science from unfolding.