Unilateral geoengineering against the global interest

A RealClimate piece on geoengineering through pumping massive amounts of sulfur dioxide in the atmosphere doesn't make it seem very attractive. While it can control temperature, it does nothing to stop ocean acidification, and I assume we end up with the Mother of All Acid Rain-damaged ecosystems on land (pretty sure that's the case but I've not seen verification). Massive pollution would have to be continued as long as humans kept adding greenhouse gases to the atmosphere and for many decades afterwards, or temperatures would quickly rebound to the highs they would otherwise have achieved. Finally there's the issue RealClimate addresses, that significant decreases in precipitation could result from this plan.

Two points I would add to this depressing picture: first, as bad as it sounds, it's not as bad as what the Lovelock-types project. While I don't think 11C warming will happen, if it did head that way, we'd choose the sulfur-geoengineering future instead.

Second, all climate scenarios, including this geoengineering scenario, have winners and losers. Right now this looks like a bad deal for everyone, but we can't predict with certainty how specific countries will be affected. Twenty years from now, some medium-sized or larger nation may figure out that its precipitation pattern won't do too badly, and so for that specific nation, geoengineering will be the best option, and then it will simply go ahead and do it.

I came up with this little problem myself, but it turns out of course that a lot of others have thought of it too. Solution's not clear other than a treaty forbidding geoengineering without international consent (along with coordinated punishment against a transgressor). I expect that would tie conservatives into knots - they'd hate a treaty, but don't want to give China a free hand.