To do this analysis and draw meaningful conclusions, there are so many variables you would need to control, including:

- population of internet users (not a lot of good data available though ITU's are ok)

- overlap of a internet user language skills with a high quality language-version of Wikipedia (maybe less important)

- cultural predisposition toward charitable giving (very difficult to quantify)

- penetration of payment systems that might work online (e.g. credit cards, paypal, direct transfer; anything but cash)

If you could control for all of those, maybe it's possible to the relationship between Wikimedia donations and GDP/capita to compare the effectiveness of messaging / fundraising in different countries.

A better way to do the broader analysis might be to use comScore data on Unique Visitors to projects in different countries (they make estimates for about 40 countries IIRC) to look at fundraising per visitor to the projects.  Of course you'd still need to control for the above factors, and it doesn't address the broader question of how well we are all reaching the broader population of internet users in each country.  You could do this analysis more broadly looking at Erik Zachte's page view data by country (e.g. http://stats.wikimedia.org/wikimedia/squids/SquidReportPageViewsPerCountryOverview.htm) but it's a bit tougher to explain when you use a Page View concept rather than a visitor.

A smaller more doable analysis might be to compare year over year performance for a single country.  So if you could get good data on internet users (or in countries where it's relatively stable now), it might be a useful way to track performance of fundraising from one year to the next.

As much as I would love to, i don't have time to play with this data now.  If anyone wants to, I did a quick data dump from comScore for January with their key data points for all the projects across the countries and regions they report.  See attached.

-s