Hi EGOL,
For sure! I also think one should monitor the branded keywords. Precisely for that reason.
However, generally the total score of my keywords (including 6 branded out of 58 keywords) will skew the graph in the competition overview in my favor... (As I figure I'd normally perform better on 10% of the total set of keywords).
So in this competition overview I guess it would be logical to filter only Non-Branded keywords, right?
(And of course our imagination and decisions make the selection of keywords limited , but I figure that doesn't represent skewed data: because those are the keywords we truly and consciously aim for... not?)