The recent snow melt theory for the gullies has used their believed pole-facing orientation to support their origin by solar heating:

NASA's Mars Odyssey Points to Melting Snow as Cause of Gullies
PRESS RELEASE
Wednesday, February 19, 2003
http://www.spaceref.com/news/viewnews.html?id=751

However, a report by Edgett et.al. to be presented at the 34th Lunar and Planetary Science Conference will argue that further examples of the gullies do not show a preference for poleward slopes:

POLAR- AND MIDDLE-LATITUDE MARTIAN GULLIES: A VIEW FROM MGS MOC AFTER 2 MARS YEARS IN THE MAPPING ORBIT. K. S. Edgett, M. C. Malin, R. M. E. Williams, and S. D. Davis, Malin
Space Science Systems, Inc., P.O. Box 910148, San Diego, California 92191-0148, USA.
Lunar and Planetary Science XXXIV (2003) 1038.pdf
http://www.lpi.usra.edu/meetings/lpsc2003/pdf/1038.pdf

The authors of this LPSC report also argue for a subsurface water model due to the observed gullies often beginning 100 meters down the slope, which they attribute to a correlation with subsurface layers.
However, in the example image they give, E11-04033, where the gullies begin 100 meters down the slope, it actually appears the gullies began quite close to if not at the top of the slope they just became more pronounced further down:

MOC narrow-angle image E11-04033
Full-resolution sample of northern wall of crater at 39.1 S 166.1 W
http://www.msss.com/moc_gallery/e07_e12/images/E11/E1104033.html

It appears that above the gullies at the top of the slope there are some incised channels that might be where the downflow of material actually began.

Also in this report the authors give a graph showing the number of gullies according to orientation. They use this to argue the gullies do not show a preference for pole-facing orientation.
However, it is notable that for both northern and southern gullies, the most common orientations are for 135 and 180 degrees away from North, that is generally toward the South. This means for the Northern gullies, the most common orientation is toward the equator, and for the Southern gullies, the most common orientation is toward the (South) pole. This in itself suggest some connection to solar insolation, which is consistent with a snow melt theory.

A key question remains unanswered in this study: is there any correlation between latitude, orientation, and slope angle for the gullies? In the paper by Mellon and Phillips, "Recent gullies on Mars and the source of liquid water", J. Geophys. Res. 106, 23165, (2001), they provide a figure that gives the maximum
temperature on a slope according to slope angle and Mars obliquity at three different latitudes. This appears in figure 5 on page 23,169. (Note that this temperature can be above 0 even
for pole facing slopes and at mid to high southern latitudes.)
Then looking for this correlation as described in the Mellon/Phillips paper would give a means of testing a correlation to surface temperature at the gully sites, and also testing the validity of the snow melt theory.

Bob Clark