It's a difficult nut to crack. How do you get across the severity and imminence without enabling the "the world didn't light on fire when you said it would!" crowd. No simple answer there.
Error bars would help. Also, defining what red versus orange means. Are the damage assessments consistent across domains, e.g. should we drop climate change as a priority to focus on novel entities?
If the purpose is to prompt policy action, I'd start by constraining the time horizon: probable damage within 10 to 20 years. Then, dollarise it. Give it error bars. Now you have a price tag an activist can talk about when convincing voters to bear a given cost. You also have a device that builds credibility--you can look back and quantify the cost of not mitigating.
But the numbers have to be defensible, or they're going to get shredded by people who don't want to do something.