Last year Dr. Bryan Bledsoe collated and presented some research on oxygen therapy, it's biochemistry, and the implications thereof. In short, oxygen is not always good, and sometimes bad. It created less controversy than I expected, at least on the interweb. (Good summaries of it all can be found here and to a lesser, but related extend here, or you can just Google for terms like oxygen and reperfusion injury to get a sense of the history).
A smallish Australian study showed up in my in-box today, adding further research to the area. While I'm not working a lot in the field at the moment (back at uni again), I think I'm seeing small and slow trends toward less O2 use with patients rolling into the local A&E. I'm curious to see what the next rounds of Clinical Practice Guidelines for the various ambulance services here have to say.