A bit late for an answer, but here is my understanding of it.
Having a two-class stock structure is a mechanism to decouple the share price from influence and eventually control over the board of directors. Class A shares, which were sold in the public offering have one vote per share, whereas Class B shares have 10 votes per share. Class B share are - at the option of the owner - convertible in Class A shares(1:1), whereas Class A shares can not be converted.

Quote:
There's been a lot of talk about people looking down on Google for having two different stock classes, the more influential of which will be held onto by Google insiders. First, why? Who cares?

Class A shareholders have less influence over strategic decisions. Some financial experts claim that the mere presence of a super-voting-power class, depresses the market value of the normal-voting-power class. It makes takeovers less likely, wether this is good or bad depends on the color of your flag. For the founders it protects innovation, experimentation and the do-no-evil approach, while it's obviously an unfortunate structure for those who would like to takeover. IIRC there are some economical theories that would accuse this system of beeing undemocratic.

Quote:
Second, what would the difference be between the Google folks hanging onto (to make up numbers) 5000 of these shares that are worth 10,000 normal shares and not having these special shares and just keeping 50,000,000 normal shares?

If the founders held the same class of shares as the average shareholder they would need more than 50% percent to be absolutely sure to keep control. With the dual class structure they are able to maintain that control even if their stake drops below 50%. Effectively it allows one to sell more shares while at the same time keeping absolute control.