Mike Carr knows perfectly well the energies and pressures associated with impacts. Unfortunately, the outburst floods don't have matching impact craters to explain their volume and flow rate. if they did, all would be simple.

An impact event might be a local trigger but it requires a regionally-pressured aquifer to already exist.

And yes, the evidence of the triggering impact would likely be swept away by such a flood, but ONLY if that pre-existing pressured aquifer were there.

See, for example, Patrick Russell's work on Lyot basin and other large impact craters in the northern p[lains which do NOT have associated flood features despite being large enough to punch through the cryosphere.

Abstract follows...

P12A-0360
TI: Probing for Groundwater in the Northern Lowlands of Mars
AU: * Russell, P S
EM: Patrick_Russell@Brown.edu
AF: Dept. Geological Sciences, Brown University, Box 1846, Providence, RI 02912 United States
AU: Head, J W
EM: James_Head_III@Brown.edu
AF: Dept. Geological Sciences, Brown University, Box 1846, Providence, RI 02912 United States

AB: Current theory on the martian hydrological cycle suggests liquid groundwater may exist beneath a confining cryosphere to depths of $>$ 10 km [Clifford, S.M., 1993, J. Geophys. Res., 98, 10973]. This framework allows geologic tests of the subsurface hydrosphere to be made based on the concept that cryosphere disruption by geologic processes should enable groundwater under hydrostatic pressure to flow to the surface. Lyot crater (215 km diameter) in the northern lowlands is large enough to have physically and thermally breached the local cryosphere in a region where low planetary elevation implies hydrostatic pressure head should be greatest. Lack of evidence of hydrologic activity at Lyot suggests one or more of the following: 1) no groundwater is present beneath the cryosphere, 2) the pore space available to water is frozen all the way through ($>$ 10 km), 3) the upper crust is not well interconnected, 4) refreezing rates are fast enough to "reseal" the cryosphere before flow to the surface occurs, 5) impacts are not effective at disrupting the cryosphere and/or providing groundwater access to the surface. Subsequent similar investigation of the ten next-largest craters (> 45 km diameter) in the northern lowlands reveals that results and implications gathered at Lyot are not due to a local anomaly but reflect conditions over ~ 20 % of the planet. Two possible exceptions are filled craters at 70 degrees N which may have been filled by 1) impact-release and freezing of groundwater, or 2) deposition of volatiles in crater cold-traps, similar to processes acting in the nearby north polar cap. While the latter is suggested by correlation of filled craters with latitude in our sample, we have undertaken further modeling to distinguish between these two processes.