Patrick Ruffini, who by his own admission is no space expert, seems to get it.

What's more striking about this accident is NASA's nonchalance, even now, in the face of the Columbia's known vulnerabilities. NASA wasn't cutting corners so much as it was accepting these imperfections as a tolerable risk. Their attitude seems to be that even attempting to fix them would have introduced other (equally hazardous) safety and engineering problems; either that, or the cost would be so astronomical as to defeat the purposes of the current Shuttle program, rendering utterly academic today's debate about whether a 5% increase here or there could have saved these seven lives. Furthermore, claiming NASA knowingly skimped on needed repairs ? and given the caliber of engineering talent working there, it would have to be knowing ? assumes that the agency didn't even have the autonomy to simply trim back a launch or two and pay for the repairs with that. Indeed, the primary alternative to the budget-cut scenario is potentially more damning: NASA knew about the tile vulnerabilities, and took a calculated risk by not fixing them.

Yup. That's life, in the non-Oprah world.

The current shuttle program costs ARE astronomical no matter what.
With astonomical costs, reliabilty will never be there, as one just cant have enough launches to have a proven safety record of acceptable levels for human transportation.