Occam's razor
If Coleman's observations were true for all the outburst flood areas, then I agree that a water-based flood model would be simple, convenient, and scientifically preferable.
Unfortunately, both Neil and myself are guilty of selective use of examples. Over the full range of flood channels, there are problems with either water or CO2 as a single explanation for all we observe.
The real world is more complex than a toss-of-the-coin either-or situation where a single paper makes a snap decision between two radical alternatives.
There is more work to be done by advocates for both water and CO2, to make a secure scientific case for the models proposed.