It's not like this was written by Peta.
In other words, I am not questioning the validity of the study at all. I suggest you read my comments a second time.
What data do you need?
Say you are flying an airplane. All of the engines fail. Would you say that "without more data that actually shows that this is a problem, it's inappropriate to suggest relighting the engines" and wait until you smash into the ground?
No. Because you can anticipate the consequences of the current situation and make changes to prevent something bad happening.
Specifically, we're noticing that there are a lot of MRSA strains on meat. So we should probably stop feeding animals antibiotics as frequently. A good way to do this is to raise them somewhere other than in a pool of their own feces, which tends to carry a lot of disease (necessitating the overuse of antibiotics).
That's where the article is going. Should we wait for a problem in case there isn't one, or should we try and prevent the problem from occurring in the first place?
Can we actually show this is causing any harm?
I'm no biologist, so I don't know whether or not we can show, for any given strain of multidrug-resistent bacteria found in a sick patient, whether its ancestors most likely acquired its resistance genes while living in a farm animal, while living in an adult human who was taking antibiotics for some other condition, or someplace else. My guess is that such a demonstration might be possible -- cheap gene sequencing, plus extensive statistics, yields a lot of deductive power -- but I don't know if it has been done already or not, and maybe I'm wrong and it's going to be impossible to do conclusively. Bacterial genomes aren't always that big, and bacteria are quite promiscuous with genetic material, so their line of family descent gets fuzzy much faster than it does with multicellular creatures like ourselves.
One thing is certain: There will be a lot more work attempting to answer these questions.