Hi Alfonso, 

thank you for comments. I'm glad that you like it. :)

For your information, the gray matter masks were thresholded at 0.99 for all cases. 
But I think I might make some errors when I uploaded the figure (or today) as I found that K-S GOF values actually already increase with CompCor regressors, and the improvement of normality by adding global signal does not seem to be worth the risk (see the attached).

The plots I created were already inspired by Power's review paper (2015, NI) and others' works. So I don't mind; instead, I'd be love to see the CONN provides such plots (with the current great GUI and performance)! It'd be also nice if you can provide summary plots (for all subjects) so that users can make an empirical decision on parameters to process their data.

Best,
-SG