The role of water in carving the Martian outflow channels has recently been challenged by the hypothesis known as White Mars. This hypothesis claims that the channels were cut by CO2 gas-supported debris flows that also resurfaced the northern plains. However, proposed analogs of cryoclastic flows are either inappropriate (i.e., submarine density flows) or are primarily depositional rather than erosional (i.e., pyroclastic flows). Subaerial mass movements on Earth do not carve long deep channels like those on Mars. I review runout efficiencies for mass movements on Earth and Mars. The efficiencies required for cryoclastic flows to resurface the northern plains of Mars are so large that they appear unattainable. White Mars seeks to resolve carbonate and floodwater paradoxes that probably do not exist. It is also doubtful whether reservoirs of CO2 could persist over geologic time in the crust. Overall, the CO2 hypothesis fails key tests and should be abandoned as a means to carve outflow channels. I present a new interpretation of the fluid source that created Aromatum Chaos and Ravi Vallis. Water, not liquid or gaseous CO2, was the causative fluid, and the source was an ice-covered impoundment in ancestral Ganges Chasma. At that time the canyon had no eastern outlet, and groundwater flowed northward to discharge at Aromatum Chaos and Shalbatana Vallis. The presence of ice-covered water bodies can help to calibrate models of volcanic-hydrologic climaxes during Hesperian time. The outflow channels, like the spectacular landforms of the Channeled Scabland, are monuments to the erosive power of catastrophic aqueous floods.