Launch Abort Options:
On thinking about Columbia, and the focus on the impact of insulation/ice from the external tank on the left wing of the orbiter on launch, I realized that there were very few options.
If the damage could have been judged accurately during ascent (and the tracking scopes plus telemetry did not indicate this at the time) the only viable options I can see are two of several available intact abort modes: Return to Landing Site (RTLS) and the Trans-Atlantic Landing (TAL). The other ascent aborts are not viable - Abort to Orbit (ATO) and Abort Once Around (AOA) would both require re-entries like Saturday's, going through the atmospheric heating of a nominal re-entry. All of these abort modes require decisions by the mission control launch officer that have very small windows, and depend on the exact timing and nature of the problem. Both of these options require a decision before main engine cut-off.
Given the apparent damage to the left wing insulation, it is likely that the normal option of preferring TAL over RTLS would have been followed, with a trans-Atlantic ballistic trajectory ending in a glide to landing in Moron-Spain, Dakar-Senegal or in Ben Guerir-Morocco about 45 minutes later. There is still atmospheric heating associated with this option, and a risk that the reheating would cause the same failure.
RTLS mode would return the orbiter to Kennedy. This decision has to be take before the +04:20 mark, otherwise there is not enough propellant to undo the kinetic energy already invested in Eastbound flight. The main engines are used after SRB separation to turn and head back to KSC. Again, the stresses on the orbiter are not small, and heating will occur as the orbiter dissipates energy on its glide slope.
Having a hole in your armour is never a good thing.