In that Grist article I talked about earlier, someone named CL Weber showed up with data:
Not looking good for grass-fed beef. Story's not completely over, though. The commenter acknowledges earlier that indirect land use changes aren't accounted for. He or she thinks that could make cattle effects worse, but not necessarily. The main alternative use I'd expect of ranchlands in the San Francisco Bay Area in lieu of cattle would be sprawl. Also it doesn't include soil carbon, which would be a significant benefit I'd expect (maybe) from switching from tilled agriculture to pasturage.
J Environ Qual 35:231-239 (2006) compared different types of suckler beef production, while Casey, J. W. and N. M. Holden. 2006. Quantification of GHG Emissions from Sucker-beef Production in Ireland. Agricultural Systems 90: 79-98 .... The JEQ article does indeed show (slightly) lower GHG emissions from organic production, at 11.2 kg CO2eq/kg-yr vs. 12.2-13 kg CO2eq/kg-yr for the conventional production schemes, which is still higher than the 10.8 kg CO2/kg-yr mean value calclulated in the Agricultural Systems piece....However, again, my point is not that grass-fed is substantially worse than grain-fed but that they're pretty much in the same ball-park GHG-wise, and that the difference between them is in the noise when compared to the difference between beef and chicken (1.4 kg CO2e/kg in the US, according to Pelletier, N. 2008. Environmental performance in the US broiler poultry sector: Life cycle energy use and greenhouse gas, ozone depleting, acidifying and eutrophying emissions. Agricultural Systems 98(2): 67-73.)
Obviously pastured beef can have much better impacts on many other things, like water and air quality, than CAFO beef, but a substantial GHG advantage can't be had without assuming large amounts of C sequestration and no land use impacts despite the large amounts of land needed.
Without those effects, beef is nearly ten times worse than chicken, which in turn is probably three times worse than vegetarian (my somewhat-eductated guess). It will take a heck of a lot to make up the difference between beef and vegetarian calories. Some interesting high-tech suggestions on page 120 of this UN FAO publication could be a start, if consumers will tolerate changing cattle's internal biota. Could be a split between the hippie and techno sides of the environmental movement.