Cool find, but can't help myself thinking that registering a domain name and submitting a paper on this to Arxiv is a bit... much. The content here could fit in a tweet or a short blog post as well. Not sure about the scientific novelty here as we're basically poking around the very top layers of someone else's software stack?