I see an interesting point in this:

How much of a difference is there between using pre-existing code in the form of slapping down some copy-pasta vs. using a known and established library? If someone has a bank of algorithms commonly used for just these types of problems, how different is using that as opposed to the real-world usage of apache-commons or boost?

I realize that copying from other competitors is wrong, but where is the line between standing on the shoulders of others and copying-as-cheating?