It's very easy to obtain 100% anonymous (even brand new, freshly minted) Bitcoins through mining.
Actually, it's not given that whatever pool you mine on is going to know at least something about you. I mined for a long time on P2Pool thinking I was anonymous till I realized anyone could see what address I was mining to and given that the IPs that are doing the mining can be seen on the network you can kind of figured out a general idea of who is mining what.
If you really truly wanted to be 100% anonymous you'd have to solo mine on TOR, right?
I guess what I'm saying is that we are seeing people spend tens of thousands of dollars on FPGA mining setups. What I contend is that if you were to buy 7300 BTC for about $37,000 today, you will be able to sell 10 BTC per day for the next two years and make more money than if you were to buy $37,000 worth of FPGA-based mining equipment and sell whatever you mine every day over the course of the same two years.
$37,000 worth of mining equipment buys you about 28 gh/s at 1.5 megawatt (FPGA's and a DD-WRT router). That's maybe 18 coins per day, enough to justify solo-mining, probably. Certainly more than the 10 BTC that you'd be selling otherwise. But over time, if other people "buy-in" to mining Bitcoin, that difficulty that gets you the 18 coins per day is going to rise, meaning which do you think is more risky, selling 18 coins per day that is bound to drop over the next two years, or selling 10 per day over the next two years.
Right now I think it is the former that is the more riskier of the two, which is why I think it is better to buy and either hold or sell over time. It would seem to me that if you think the price or difficulty is going to be increasing for a while, you should be buying/holding and if you think the price or difficulty is going to be decreasing for a while, you should be selling/mining.