I've been working through the paper for almost two hours now. I'm not an expert on static analysis, so I don't know how much of the information in here is novel

I've been working through the paper for almost two hours now. I'm not an expert on static analysis, so I don't know how much of the information in here is novel, but the paper does a great job of concisely conveying a lot of interesting ideas. His model for representing data exposures is beautiful in its simplicity, and I think it could almost trivially be adapted for dynamic analysis too. Definitely worth a skim in my opinion.