Thanks Flight... I'll try and be brief.
Meta-analyses are really only useful for indicating future areas of properly-controlled hypothesis-based research; not as suddenly revealing answers to questions that could not be found in the original studies.
The data on which this meta-analysis is poor, as admitted by the authors:
"...study quality was generally moderate, with only one of seven included studies rated free from bias, and the quality of the others was limited by poor reporting. Thirdly, studies varied both in the outcomes measured and the consistency of reporting of results, which severely hampered our ability to aggregate results."
I would encourage people to read the freely-available full article, because apart from the dense statistical table section the rest of it is quite easy to understand. Especially the limitations.
Lastly, all the studies included were on seasonal flu, not swine flu, and the risk/benefit of giving antivirals is quite different for both types of infection.
The Andrew Castle thing is a good example of how confused people can be about the side-effects of oseltamivir. It is not associated with precipitation of asthma attacks, whereas swine flu is. So rather unfair to complain that it was the medication that was at fault, when it could have been an infection itself.