In a post written in late February, Dan Trombly raised a number of practical, logistical, and strategic issues which, at a minimum, would severely complicate any effort on the part of the U.S. to arm the Syrian rebels in a way that ensures that the weapons reach the non-jihadist sectors of the opposition and in a way that contributes to the emergence of a post-war situation in which the U.S. would have any influence even over the factions it had armed.
Though characteristically sharp in its analysis, Trombly's post does not mention explicitly what should be the single most pressing issue for all countries, including the U.S. and Britain (both of which have recently said they will send non-lethal military aid to the rebels): namely, how best to reduce the total amount of suffering? More than a million refugees, and that counts only those who have registered with the UN, have fled Syria for Lebanon, Jordan, and Turkey. Humanitarian aid has been pledged but most of it has not yet been delivered. The refugee situation is threatening to become a full-scale catastrophe. (See, e.g., the report contained in this segment.) Moreover, the civil war has already taken roughly 70,000 lives and as both sides cast away whatever minimal restraints they have hitherto observed and as the Assad regime uses missiles indiscriminately, one can expect the casualties to continue mounting.
In this context, the right question is not how the U.S. can best protect and advance its narrowly defined geopolitical and strategic interests. Rather, the right question is how the U.S. can best help minimize the duration and the total amount of human suffering that the current situation has produced and seems very likely to continue to produce. I don't know the answer to that question. But even the most sophisticated strategic analysis, if it focuses solely on the narrow question of 'interest', will end up being less helpful and illuminating than it could be.