From what I can see, the results of this study don't prove (nor are they claimed to prove) that excess selenium causes diabetes. There's an association in a relatively small group of people that may not be borne out in more comprehensive studies.
I also don't see any convincing evidence that selenium supplementation prevents cancer, though trials underway may answer this question more definitively.
The media have a bad habit of jumping on preliminary reports as though they were conclusive, when they are not. Some skeptics will gleefully cite this study to dismiss selenium supplementation. Others will darkly hint of nonexistent Conspiracy theories by Big Pharma (which would love to sell you all the supplements it can).
Science can be dangerous when swallowed whole in small doses.