Good effort, but personally, this should not have a sticky.
Few reasons: you did not create the program.
That means, you don't know how effective it is, its accuracy, or what it determines to describe a "character", or how it achieves its algorithms.
For example, are the characters counted only when they are logged in? What if a character never logs in, but has been created?
What if the person has a character as an alt? Does it determine has one unique character. Does multiple characters per account be classify as one character?
Is there a period of time it has to check? I assume it does not go into the database, so it checks the online players from a certain period. That is, there may be a lot more players/characters, that has not been checked.
Census of classes, so what? What do those statistics say? Within the last 1 week? 2 weeks? 3 weeks? These classes/races have logged on?
The amount of ambiguous data makes this difficult to interpret and extrapolate effective data since these algorithms seem unreliable.
This is just a small subset of the problems available with this program. However, a good attempt of 'approximating' the amount of players. Since it's not an official source, and far from accurate, then I disagree with adding a sticky.
A sticky implies it is a good, useful, and accurate source, which you cannot prove how these algorithms work that are effective and accurate to the extent that works as intended. The best data source would be from Warmane's database, which I doubt they would provide.
Hope this helps.
Baleroc