IMO there isn't much discussion necessary. All the information is already outlined on Wikipedia articles.

Democracy is the system where a number of people (in this case the eligeble people are outlined by Nokia) elect from a list of candidates (in this case selected by a rule as outlined by Nokia); the winners will represent the people who are under their reign (NOT only those who voted for you; seems to be a missconception...). Between these 2 groups of voters and candidates we want the procedure to be as democratic as possible so that the people who choose are the best represented. That is the goal of democracy which current governments are by far not able to reach because of insincere and strategic voting. Democracy is not a 1 or a 0 despite what some might want you to believe.

On the Internet we have the opportunity to implement a better system than Pluratity. Lets make use of this.

If you want to touch on the subject of complexity, by all means, quote me on Condorcet. Its the only reasonable argument against Condorcet I can think of. Approval is not complex at all compared to Plurarity.

Example:
lardman: YES
Karel Jansens: YES
Reggie: YES
allnameswereout: ABSTAIN

Means lardman, Karel Jansens, Reggie get +1 and allnameswereout gets +0. Count 'em all up, highest score wins. I'm pretty sure the entire world is able to say yes or no when asked. With Approval, in the US, the Democrats and Republicans might actually gain some competition during the elections. Competition is good...

Condorcet is much more democratic. Its used by communities such as Debian for long time now. Open source tools to calculate the outcome are available. You can find all kind of examples on Condorcet, preferably with SSD as underlying ruleset to solve the maths. They're using a CLI utility which outputs the mathematics which are then posted on mailing list but IIRC theres even GUIs available.

Heres some Condorcet insights and compares http://rangevoting.org/EMorg/indx.html