A while ago I talked a bit about getting some key data done for free. Well, that didn't turn out so well. We did eventually get the data, but they were about a week late for our grant deadline and it turned out that the Data Producers left out one really essential step that meant we got data of far less use than we had anticipated given the instructions we sent. But again, free.
So after an initial scan of those data and a limited amount of time playing with them, I put them on the back-burner for a while because of teaching and other commitments. Then we got a new toy and all of a sudden these data can be manipulated in a very different way.
It turns out that, while not entirely useful for the question we had in mind, the data are incredibly useful for providing some proof of concept stuff for a proposal that is in the works. I never would have done it this way given the choice, but with a little thought and an enormous amount of CPU time, we may actually be making lemonade out of crap. Mmmmmm.