This is a really cool idea! I wish that they had spent a little more time examining non-turbulent images; really, once the analysis software is written, I'd ex

This is a really cool idea! I wish that they had spent a little more time examining non-turbulent images; really, once the analysis software is written, I'd expect to see all of van Gogh's work piped through it, and a handful of other paintings, both impressionist and not, all graded on turbulence. I'd like to see an actual numerical comparison to the gaussians, rather than letting us guess at "yeah, that one looks pretty close". One non-turbulent image for comparison (classified as non-turbulent "as it can be seen") is not really enough. And there's no link to the code! When are researchers going to understand that they haven't exposed their process to review until they've given reviewers access to the code?