I've not read the paper but am impressed by their presentation and the experience and community which they have already created around their 'SafeNet' browser.

I've not read the paper but am impressed by their presentation and the experience and community which they have already created around their 'SafeNet' browser. I'm inclined to believe their mathematical analysis of their own system is sound. They say its based on a system with proven O(n^2) messages requirement, with an elaboration reducing message requirement to a more scalable O(n log n)