Other than the band, do we keep track of dna or some other bio marker to make sure someone didn't just move band on some other Albatross?
Also keep in mind that these are wild animals, so life expectancy rules of stone age humans would be the only thing similar (i.e high child mortality, minor injuries can kill you etc.).
Finally, the article also mentions that her tag has been replaced six times. So, there may be older tagged birds. But if they lose their tag once, they become young animals again. Which would skew any statistic.
How important these tags are for estimating bird life expectancy? If they are, isn't this going to cause, like, typically updating the Bayesian priors towards less longevity than they actually have? I'm not working in this field at all, so I'm guessing there are methods in use to counter this?