There is an interesting story in today's New York Times about tile damage on a previous Columbia mission. I think Danny discourages full text pasting of copyrighted text, so the link is here. (Registration required, but it is free.)
http://www.nytimes.com/2003/02/04/national/04WRON.html
The story relates a report on tile damage to Columbia in 1997, in which the number of tile hits greatly exceeded the norm. The tile damage was caused by foam falling off the external tank.
"Inspectors counted 308 hits. Of those, 132 were "greater than one inch." Some of the hits measured up to 15 inches long with depths of up to one-and-a-half inches. The tiles [presumably the ones hit] were only two inches deep, [some tiles are five inches thick] so the largest hits penetrated three-quarters of the way into the tiles." [Comment, if these tiles were only two inches thick, these were not located in an area of high aerodynamic heating.]
The damaged tiles were mostly around the shuttle's nose. After the mission, more than 100 tiles were taken off because "they were irreparable."
NASA also changed the formulation of the foam after the 1997 incident. The Times article states that "... to be environmentally friendly, NASA had eliminated the use of Freon in foam production. The Marshall Space Flight Center in Huntsville, Ala., later concluded that the absence of Freon led to the detachment of the foam. ...the formulation was later improved."
Finally, the Times article includes opinions from several aerdynamicists involved in shuttle aerdoynamic testing, who said 'that even slightly damaged tiles — perhaps only roughened or cracked — could generate turbulence near the tiles during the tremendous speeds of re-entry, creating potentially dangerous heating of Columbia's aluminum skin.'"
_____________________________________
Over 30 years ago when the shuttle was first proposed, the scheme (and corresponding launch rate) was for it to be mainly used as a logistics vehicle to transport hydrogen for off-loading to in-orbit NERVA rockets that would be used for manned lunar bases and manned Mars flights. NERVA rockets used a nuclear reactor to heat hydrogen as the means of propulsion; ultimately, work stopped on these engines just prior to final flight test.
Initial shuttle designs called for two air-breathing engines so it would not be a glider on return (the USAF was most interested in having an ability to land at points some distance away from the orbital track) and for the booster rocket to also be manned and return from a sub-orbital flight to a landing site.
The lunar base and manned Mars exploration missions were much like what Arthur Clarke had whirling around in space in "2001, a Space Odyssey". The basic hardware for these missions would be launched into earth orbit by the bottom stages of a Saturn V type rocket, there to be assembled by astronauts. Use of a Satrun V meant you were not constrained by the weight-to-orbit capabilitiy of a shuttle, nor by its payload bay dimensions.
But in the words of the refrain written by Kurt Vonnegut (who was a prisoner of war in Dresden when it was fire-bombed) in his novel "Slaughterhouse 5": "And so it goes."