Thanks for bringing that up. I've actually argued the opposite in the case of voting. Using timeless decision theory, you can justify voting (even without causing a bunch of people to go along with you) on the grounds that, if you would make this decision, the like-minded would reason the same way. (See my post "real-world newcomb-like problems".)

I think a crucial difference between the two cases is that non-pollution makes it even more profitable for others to pollute, which would make collective non-pollution (in the absence of a collective agreement) an unstable node. (For example, using less oil bids down the price and extends the scope of profitable uses.)