GMs vs. 2100-2200 (II)

In GMs vs. 2100-2200 (I), I noted that the GMs were more consistent in their choice of first moves (see that original post for the background of my 'experiment') and wondered if it was a random result. I repeated the experiment on the games from the top ten boards of the first round of the 2008 Mainz Chess960 Open and produced the following chart.

SP884: RBBQKRNN

It again appears that the GMs were more consistent in their choice of first move. Four of the five games with the GMs as White started 1.c4, which was not chosen by any of the lower rated players as White. Are the GMs using some kind of chess logic which escapes the others?

[It should be mentioned that the software I'm using to produce the diagram doesn't handle chess960 castling. That's why some variations end as soon as one side has castled.]