You are very right: Facebook (and Twitter, etc) are already being targeted by this. We're currently trying to quantify the scope of the problem on Facebook.
Stopping crowdturfing is an extremely hard problem. Traditional anti-spam techniques all assume a certain amount of automation on the part of attackers. Mass e-mail spam can be detected by using statistical methods and machine-learning to assess content similarity and look for templates. CAPTCHAs and other Turing tests can hinder bots from logging into social networks, and thus quash the flow of spam. However, when spam is generated by humans, these assumptions are totally shot. Turing tests don't work, each piece of spam content can be tailored to be unique, etc.
Legislation will probably need to be part of the solution, but it won't be 100% effective. Currently, crowdturfing sites operate in the open. Criminalizing them will push them underground, which will hopefully reduce their attractiveness to workers (and thus reduce the amount of spam that is generated). The ban would need to be global though (good luck with that...), otherwise crowdturfing operations can just move offshore.
