I suspect this to be in the genre of the ask-for-something-impossible-on-rentacoder-and-see-what-happens experiments. Don't PhD candidates have to get their proposals accepted by their advisors and then have periodic reviews during the process?