Thanks for the link, ORAC
The article you link to says:
-------- Start quote ---------
… Even with replacements over the years, 70 percent of Columbia's tiles were the originals… Surface roughness ……. has to do with the gaps between the shuttle tiles and the "step," or difference in height, between one tile and its neighbor. …….The roughness of the wing is indicated by a measurement called the K equivalent, derived by combining data on the gaps and steps with information on the airflow …
-------- End quote ---------
The impression I get now is that the roughness is a left-over of the tile-installation learning curve from 20+ years ago. In that case, there may be nothing wrong with either the tiles or the wing, just that the installation methods had not yet been developed sufficiently when the first Shuttle was built. I wonder if the left wing was built first?
One might guess that if tiles projected various distances due to imperfect installation, then the adhesion in some cases might not have been too good either.