I of course concur that conflicts arise, and resolutions may be painful or even arbitrary.
However one assumes rightly or wrongly that the state takes all factors into account before legislating such that the chances of "uninformed consensus" in minimized. Thus your arguments about "obligation" and statusof "conensus"are theoretically avoided.
I have no idea why my arguments about "obligation" and the status of "consensus" are theoretically avoided. I'm not even sure why you'd put "consensus" in quotation marks. You, after all, are the one who came up with the idea of consensuses in the first place. If you're not sure what the nature of a consensus is, then maybe you should reconsider your entire position.
If a social consensus is not obligatory on the members of a society, then certainly a social consensus on homosexual marriage amounts to nothing more than a suggestion. Indeed, it's hardly worth mentioning. On the other hand, if it's obligatory, then there's no sense in saying that the question of obligation is theoretically avoided: far from being avoided, it would be inherent in the very notion of "consensus."
So, you have your choice: either the social consensus that you posit is, at most, precatory, or else it is obligatory. If the former, then any social consensus on homosexual marriage is largely irrelevant and your reliance on it is unwarranted. If the latter, then there is the chance that one obligatory consensus will come into conflict with another. And if that's the case, what happens when one consensus contradicts another?