What surprises me is that this contingency was not already in place prior to the Columbia accident. Was it budget constraints or did they really think it was that unlikely that a shuttle would be rendered unable to return from orbit and need the crew rescuing?
I think more likely, at some point, you just have to accept the risk.
How far do you take it? 4 launch vehicles all on standby ready to rescue the one before it? Is 3 enough? Maybe 2? Up until then, they had obviously calculated and accepted the risk of having no standby.
So obviously before the incident they were accepting the risk of no standby, where-as they revised that and decided that for a particular mission where using the ISS as a life-boat was not an option, they needed a standby to accept the risk.
