Jorge R. Frank wrote:

or for the crew to power down and try to hold out until Atlantis could be
launched to rescue them (which was perhaps just barely possible, but would
have involved risking a second orbiter and crew to the same problem that
doomed the first).

We've all read enough about Columbia, but this point always seems to go
unchallenged. Knowledge of a new failure mode doesn't suddenly make it
more likely. However you slice the data the chances of Atlantis suffering
the same problem were very slim, and I'm quite sure that there would have
been no shortage of volunteers willing to bet against it. And whatever
system you're using, every launch and every re-entry carries some risk.

As has been pointed out, the far greater risk to Atlanta would have been
from rushing the job.

Still wouldn't have happened tho.

--
A man can always find a pub.