The final two paragraphs of the paper (with references omitted) are instructive:
QUOTEThe ice age scenario described here provides a comprehensive picture of how global scale glaciations may have evolved on Mars. It predicts that the top surface layer in the mid-latitudes has been reworked numerous times during the past few million years, that much of the ice seen by GRS is diffusively formed pore-ice and less than half a million years old, but that an older ice sheet exists at higher latitudes. The pore-ice layers extend and retreat close to synchronously on both hemispheres. Over the last 2.5 million years the original ice sheets receded vertically by an estimated 60 cm (creating less than that in additional sublimation lag), considering only areas where it remains within the topmost metre today. Most of the ice loss, however, occurs at the retreating edge of the ice that lies below one metre, and the total ice volume change may be as high as 10^5 km^3, an amount that is still small compared to the volume of the permanent polar deposits. These movements are probably the counterpart of the layers observed in polar regions.
The dynamic nature of the ice sheets makes Mars an ideal system in which to test and expand our knowledge of astronomical climate forcing. A great deal could be learned about terrestrial ice ages from the study of martian ice stratigraphy—a longer, cleaner and simpler record than Earth's.