Having thrown out the use of Wikipedia markup parsing libraries I was able to focus on just getting what I need to work to work.
Good progress so far - I've been able to parse some good chunks of the content I want. With the added bonus that I don't just get 2015 results but instead all results that are available.
The only hiccup is that there are effectively two page templates I need to parse, and I've only done one. Hopefully the other shouldn't be too hard though.
Then I need to go back and check the results. Still trying to think what the best way of doing that will be.
[Image credit: Sean MacEntee]