Pannel discussion on Bitcoin with Roger Ver, Tony Gallippi, Shakil Kahn at LeWeb 2013, one of the most influential tech conferences in the world.
For the question at the end about the "datacenters owning the bitcoins", a similar argument is how transaction fees go to the miners so the miners will be the ones who end up with the bitcoins.
That isn't a problem though because mining has a trivially low barrier to entry, such as this ASICMINER USB block erupter:
and thus there isn't much seigniorage over the long run.
This is in contrast to fiat banking which is a regulated monopoly, consuming hundreds of billions of dollars of fees just in the U.S. (e.g., NSF fees, account maintenance fees, etc.), merchant fees (paid by the merchant but whose costs are built into the price charged), etc. With the banking system there is a structural flow of funds from the customers (user) to the provider (banks). With bitcoin, that's just not the case.
As far as the miners having control of the network, that's not really true. The miners are essentially employees. Tony described this perfectly at 30:50 in the video regarding the March 11 hard fork. Hey said "which [side of the fork] which the majority of the users" is what determined which side the miners switched over to. And that's a perfect example of where miners didn't call the shots, ... nor did the Bitcoin developers. The mining capacity went to the side where those who pay the miners were on. Because Mt. Gox is still such a huge percentage of the market, if they had been on v0.8 that side of the blockchain fork would probably have remained the longest chain and everybody else still on v0.7 would have needed to take emergency measures to accommodate the resulting locking issue. i.e, the economic majority (those who have coins and those who buy coins) are the ones who control the miners and not the opposite.