The anti-inflammatory effect of statins is indeed well documented. Looking at the NNT, however, makes me skeptical about the effect size.

I have serious doubts about CRP replacing anything in the toolbox of risk prediction. The reason is simply that it doesn't add more than marginally to the predictive value of the established markers (of which calendar age is still the elephant in the room). And on its own, the sensitivity and specificity of CRP are too low to rely on.

I also don't think that statins are replaceable with PDE5 inhibitors, for the simple reason, that each drug class targets different pathways. Lowering cholesterol is one thing, relaxing arterial smooth muscle is quite another. As you said, lifestyle habits like exercise and diet are still the best options.