Typically in this sort of analysis one attempts to weigh more recent statistics more. I'm not even sure how you got to .270 given that that's the maximum of their career averages. Assuming you broke it down in splits, you completely neglected the fact that there are many more RHers than LHers in the league, and so you should weigh the stats towards the .250 of Snider, who would be playing more.
Using recent statistics more, Rajai hasn't hit his career average vs LHers in a few years, and Snider has hovered around his. But, to remove some of the stupidity of this argument, one should also note over the last several years, the OBP of Melky has been much better than both Snider and Rajai, and if we use statistics from the last 3 years, which one could reasonably argue are much more indicative compared to their career rates, Melky is simply an incomparably better than a Snider/Rajai platoon, given Snider would get many more ABs.