Agreed. His experience and transformation is interesting, but the article is pretty lacking in details.
> Searching for "site:arxiv.org Jason Padgett" results are filled by works of "Deborah Padgett, Jason A. Surace" and such,
Small world! Jason Surace and I are both members of a collaboration studying galaxies which are forming stars a rates 10s to hundreds higher than "normal" galaxies (like our own Milky Way). Just amused to see his name outside of the normal context in which I see it. :)
We are both members of the "Great Observatories All-sky LIRG Survey" (GOALS; http://goals.ipac.caltech.edu/). It is a survey of 200 galaxies in the nearby Universe (within ~1 billion light-years) which all have enhanced star formation rates. In this case, enhanced means they are forming 10s to hundreds of solar masses worth of stars per year (compare to our galaxy, which forms ~1 solar mass worth of stars each year). These galaxies are LIRGs (Luminous Infrared Galaxies), which means they were selected by the fact that they have emit a significant amount of their total energy in the infrared. For most of these, that infrared luminosity comes from dust, which has absorbed and re-emitted the optical and ultraviolet light from the young stars which have recently formed (in the past 10-100 Myr).
When you look at LIRGs, say with the Hubble Space Telescope (http://www.spacetelescope.org/news/heic0810/), you see that many of them are galaxies in the process of merging with other galaxies. My focus is on matching computer simulations to individual merger systems. Since these mergers can take up to a billion years to go from start to finish, simulations enable us to "fast forward" and "rewind" each merger. One research interest it to use the simulations to determine how differences in each system's encouter correlates with how much their star formation is enhanced. I'm still working on it so I don't yet have the answer, but hopefully in a couple years!
What's about the rest of your post… You see, I don't think it's about 'mathematical savants' or anything. And, once again, this guy doesn't draw fractals. Anyway, for me it isn't important part about that. "Genius" is just a bold word after all, it means very little if anything at all. Somebody achieved something, or he hasn't, that's all. As it seems to this point with information we have this guy isn't anything but somebody who behaves oddly after damaging his head, which isn't big deal as I believe anybody would agree. Yet if he's loud enough, journals are taking interviews with him, and links to these interviews are posted on HN, and people discuss that magical phenomena with excitement, and he writes a book, and somebody probably buys this book, and somebody with degree in neuroscience writes dubious paper about that and people call it scientifically proven, and some other journalist writes long but pointless article about that, and so on, so on, so on… And even if there was actually something interesting about his case, it's pretty hard to see it now, after all that "mathematical marvel" screaming, especially when no real facts given. I find it sad and much more important than if he is actually a "marvel" or he isn't.
Anyway, something even worse: searching for "Jason Padgett" shows many results on some journal sites, facebook, amazon, fineartamerica.com, whatever. And guess what? Nothing on arxiv.org! Searching for "site:arxiv.org Jason Padgett" results are filled by works of "Deborah Padgett, Jason A. Surace" and such, and no "Jason Padgett" on the first 2 pages! Weird, huh.