A lot of hot air has been spewed lately over this issue.  I offer you
the following simple graph:

http://www.aliasimages.com/Thirds.jpg

The findings (not that this is new by a looooong shot):

2/3 is not the golden mean!
                ** er, yeah, we knew that.

Any subject (bigger than a point) placed at the golden mean or the
two-thirds point is not likely to be distinguishable from the other in
any important sense.
        ** no sh.t Sherlock.

"It takes a little sophistication to understand why 1.618 works and the
Rule of Thirds does not."
                ** No comment.  Anyone who takes themselves
                   that seriously, but has never produced a
                   photo worth comment, should really find
                   something better to write about.

Cheers,
Alan

PS: In closing we do invite Tony Polson to back up his claim of having a
cover photo on Paris Match.  Or anything at all worthwhile from his
shooting pace of 50 rolls per average *week*.