And we're back on topic! (Vaguely. At least it's computing-related.)
Bitcoin is a bad idea, for one reason that has nothing to do with its bizarre hyper-deflationary properties or its pyramid-scheme-like massive privileging of early adopters (who, in a world where Bitcoin becomes successful, do their mining before the supply of currency chokes off and everyone has to start using fractional Bitcoins for everything, the opposite of what happens to early adopters in most currency schemes and an excellent reason for late adopters to be suspicious of it).
It's a bad idea simply because it is utterly dependent on networked computers run by individuals remaining secure: an attacker can trivially steal your keys, and all Bitcoins controlled by them, and as so often with computer security, they only have to be lucky once. Worse yet, unlike with many other consequences of successful system compromises, restoring from backup and fixing the holes won't help you: the money is gone into your attacker's pockets. That this has happened repeatedly, even to major Bitcoin exchanges, suggests that the scheme is unusable by anyone who cares about the security of their money until such time as computer security improves radically.
Bank systems are a similar target in the real world, but at least there they have a non-computer-security-dependent monetary system to depend upon, and have legal guarantees and the like to ensure that if you are defrauded or the bank collapses you will be reimbursed, at least to sufficient degree as not to pauperize you.
I am rather surprised that, post-2008, anyone would consider a non-governmental monetary system worth the risk. Haven't depositor guarantee schemes proved their worth in the last few years? It's not even useful if you want to hide your transactions from the government: it's *much* more trackable than real-world money, and is trackable forever, though tying accounts to humans may be hard.