streaming feedback on final (no action necessary)

[jovo]

• baby is super cute
• what if you do only the high amplitude shrinkage, and skip the wavelet stuff?
• why do the distance matrices look less blocky than before? just more subjects?
• what happened to perfect discriminability?
• if task time-course is similar across subjects, then if we remove it, we might be removing task-based signal.
• perhaps optimize on task discriminability rather than subject
• what is the distance used in discriminability
• a table comparing to EEGlab explicitly would be nice.