> Published in: Volume 33 Issue 4, July 2014 
    > Software - Code available soon.

The tragedy of modern science, where scientists care more about getting published than enabling actual scientific progress by allowing others to reproduce their results and continue work building on these results.

I really hope that scientific publication platforms will begin refusing papers provided without proper reproduction means before more research is practically lost.

You could try emailing the authors?

Fairly obvious extension: when arrived at a composite image, choose the one image from the set that is closest to it and show that instead of the composite to get a sharp version.

It's not really new. I've seen such work back in 2007 (http://boston.conman.org/2007/08/28.1) and an update in 2013 (http://boston.conman.org/2013/01/03.1). Sure, the examples I saw were geared towards human faces taken at a particular angle (face on, head filling the frame) but the concept is the same.

With enough data, almost everything becomes interesting and a path to new understanding.