The O2 absorption comes (as a previous poster indicated) from an easily oxidized material, such as iron, present in or under the membrane (Cichello, 2010). To suggest that it will strip O2 from CO2 ignores science. Then again we've only known about yeast's role in fermentation for 156 of the past 3,800 years, so superstition plays a large role in brewing it would seem.
Having judged a good many beers, I can attest to the strong correlation between large amounts of headspace and oxidized taste profiles. It is possible the lack of care exhibited by a sloppy fill is the precursor to the off taste and not the headspace itself, but in my experience a careful brewer makes a lot less bad beer than a sloppy one.
I have tasted a good number of beers capped with these and I have not noted any metallic notes which really would have been my only concern. If it makes people feel better, great. It does no harm, may do some good, and that's a lot better than you can say about some of the gimmicks out there. The science is there for the examination in the food packaging industry so I don't see why it's such a stretch to make the case (excuse the pun) for using it in beer packaging.
Cichello, Simon Ph.D. (2012). "A Guide to Oxygen Absorbers". Wholesale Group International Pty Ltd. Retrieved February 9, 2013 from OxygenAbsorbers.net.