Thanks for the links.

Ivanov and Head note that there are other explanations for the ridges observed by Withers and Neumann.

Ridges are also known to appear on the sea floor. See Figure E on the page:

Coastal Processes and Landforms
"E. Seafloor sediments are being piled up in ridges along the Oregon margin like folds in a carpet as they are scrapped off the Juan de Fuca Plate by the over riding North American Plate in a process called subduction."
http://geomorphology.geo.arizona.edu//geos450/LECTURE22/lecture22.html

Also such ridges have been described as actually eskers:

Evidence for an ice sheet/frozen lake in Utopia Planitia, Mars
http://adsabs.harvard.edu/cgi-bin/bib_query?1993mnps.work....4C [abstract]

Geomorphic processes in the Argyre-Dorsa Argentea region of Mars
http://adsabs.harvard.edu/cgi-bin/bib_query?1993LPI....24..753K [full text]

GIANT PALEO-ESKERS OF MAURITANIA - ANALOGS FOR MARTIAN ESKER-LIKE LANDFORMS
http://www.lpi.usra.edu/meetings/polar2000/pdf/4031.pdf [full text]

Or also subglacial or submarine hyaloclastic ridges:

CHARACTERISTICS OF HYALOCLASTITE RIDGES UNDER WESTERN VATNAJÖKULL
http://www.lpi.usra.edu/meetings/polar2000/pdf/4062.pdf [full text]

Perhaps it can be determined if the ridges they observed are preferentially found near glacial or submarine landforms.

Also, Withers and Neumann did not consider the landforms in Deuteronilus Mensae, a region considered to show the strongest evidence of a prior ocean, such as high scarps. Deuteronilus Mensae also separates the northern plains from the southern highlands.

A report at the Mars Polar 2000 conference ascribed landslides seen in Deuteronilus Mensae as caused by glacial ice. However, another possible explanation is submarine seafloor collapse. See Fig B on this page:

Coastal Processes and Landforms
"B. Groundwater seeping from low-lying Florida weakens the deep seafloor off the West Florida coast, periodically causing locations to collapse, a process that over time has formed a sheer, mile-high cliff."
http://geomorphology.geo.arizona.edu//geos450/LECTURE22/lecture22.html

In the Malin and Edgett paper that failed to find definitive evidence of ocean related landforms in MOC images, note that one such landform sand dunes are short lived at geological time scales anyway. In fact Malin and Edgett used this fact to argue in their later observation of Martian gullies that they were recently carved:

Martian leaks: Hints of present-day water
Science News, Week of July 1, 2000; Vol. 158, No. 1
"Four lines of evidence point to the youthfulness of the features, he notes. None of the gullies are dimpled with craters, indicating they are not old enough to have been bombarded by space debris.
"In one of the images, a gully and channel lie on a field of sand dunes. "Sand dunes tend to be very young features," and the channel that slices through them must be even younger, Edgett says."
http://www.sciencenews.org/20000701/fob2.asp

I found a web page that discussed the Malin and Edgett work on lack of ocean landforms in MOC images:

MAINSTREAM MARTIAN WATER & GEOLOGICAL DATA
http://www.albany.net/~rwcecot/iraap/Mason/Mars.htm

Mason's is an anomalist site. However, he makes a couple of interesting points. First he suggests rotating one of the images that Malin and Edgett examine so that craters will have the appropriate relief. He notes that M and E imply that the image shows the supposed shore is higher than the plains of the ocean. However, rotating the image shows that the shore does have higher relief than the ocean area. There is not a high cliff separating the areas but one could suppose this is due to it being a shallow lake or sea.
Second, he gives examples of areas in Australia that are known to have been ancient shorelines yet it is still difficult to make out definitive ocean landforms from orbital imagery.

Bob Clark