A study of Rutgers University students testing their automatic attitude preferences for environmental politicians versus anti-tax politicians found a significant shift to the environmental candidate after Hurricanes Irene and Sandy (full article behind paywall). Physical evidence literally hitting you in the face may satisfy the information deficit. Hopefully someone can check on Colorado in a little while.
I'm pretty certain that in 50 years, the number of climate deniers will be similar to the number of Flat Earthers no matter how well or poorly we communicate the issue. Objective reality has a role to play. The issue is how much sooner than 50 years from now we can get people to take required action.
Following up on Eli's post on the Kahan paper, I have a thought experiment to follow up Kahan's study: what if immediately after the subjects had completed the study, the researchers explained to them how the math actually works and then asked them if they wanted to revise their answers? It seems highly likely there would be a tremendous shift to the correct answer (and if the low numeracy people didn't shift, that tends to support Michael Tobis' view). This result would support the information deficit model. It's questionable how closely this resembles what happens in the real world, but the same could be said about Kahan's setup.
Obviously framing and psychological identity play a role in getting us to confront climate change soon, but so does the science, and so does the exposure of bad arguments used to deny the science.