A lot has been said about the "safety culture" at NASA. The best way to accomplish that culture would be to scrap the whole shuttle program and start over, especially after discovery that the problem that doomed Columbia has not been fixed.
The key problem is that there are elements to the launch system that sit forward of exposed critical elements. After recent experiences, it is clear that the whole side-by-side configuration of orbiter and expendable boosters will continue to cause problems.
From a materials standpoint, I do not see how a reliable fix can be created. The insulation surrounding the liquid oxygen main booster will be under extreme stress during lift-off. Add to that the inevitable vibration, even from the sound of the engines transmitted through the body, and you have a high-cycle fatigue system with an elevated average load. Any compound used as an adhesive will be going through hell. Given the large surface area of the main booster, it would be unfeasibly difficult to maintain the perfect bonding necessary to prevent stress multiplying bubbles that would further throw the bonding system further off optimum.
From the safety perspective, it would be far better to have one of three systems.
1. Single-stage-to-orbit: Far less possibility of debris impacting critical structures simply because there are fewer elements that shed. There would be a possibility of a forward element shedding and striking the aft portions, but it would be more likely that any spalled pieces would travel outward rather than directly back.
2. Critical structures forward. This is the classic rocket configuration, the payload is forward of the boosters and any elements they might lose in launch. Not very useful for something the size of the shuttle as the resulting structure would be gargantuan.
3. Low stress booster system. Essentially similar to the White Knight/Spaceship One configuration, the primary lift to elevation happens slowly, White Knight took an hour to climb to 50,000, and separation occurs prior to primary ignition. Lower stress on the "booster" translates to an easier engineering problem of keeping all the bits contained.
As always, feedback and corrections are appreciated.
Update: While one can always find someone who has at some point in the past complained about a particular flaw, the whole idea of degrading the performance of critical applications over feel good public relations "environmental concerns" should have been setting off warning bells. The old foam used to insulate the main tank was originally made with freon.
NASA's limited insight into changes vendors had made with materials used in making the tanks.
Environmental requirements requiring removal of freon from the process for spraying the foam insulation onto the tank. NASA has said that the freon-free application method resulted in foam that initially did not adhere to the tank as well, but changes were later made to strengthen the bond of the environmentally friendly foam.
Time for NASA to check their priorities. No engineer worth a damn should be willing to trade of safety for some public relations gimmick. The only environmental issues that go into the shuttle should be the environment the astronauts have to work in.