On that post on the non-random nature of natural selection, Keith says: "evolution cannot prepare them for some future change in that environment." Remind me how this applies/interacts with regards to preadaptations.
That term preadaptations does imply some element of advance preparation, doesn't it? I actually prefer the word exaptations, because it's not so loaded. An exaptation is a feature that has adaptive significance (ie it's of survival value) in the organism's existing environment, but which turns out to be advantageous in a new environment as well. However, the fact that a feature may have value in a changed environment doesn't override the fact that it must be of survival value in the current environment to have evolved at all.
This site has an excellent discussion of exaptations in the context of the evolution of insect flight. Another example would be the evolultion of tetrapod limbs. Acanthostega is an early aquatic tetrapod which had limbs – these can be viewed as exaptations: they had survival value in the aquatic environment, but would also be of survival value when the early amphibians first ventured out onto the land.