I'm pretty skeptical of Bitcoin's chances at this point.  Not because the underlying technology isn't sound, but because it's been completely co-opted by speculators and "the USD is just, like, paper man!" doomsday nutballs.  I think the problem lies with the mining algorithm:  it's a great idea to use mining to incentivize peer to peer participation in the payment processing network, but the model of exponentially-increasing difficulty is just foolish.  A model where participants earn a share of transaction fees based on the number of actual transactions they help process would be far less wasteful of electricity and CPU cycles and could potentially encourage far broader adoption of the Bitcoin software, thus paving the way for it to be used as an electronic currency by more than a handful of enthusiasts.  There wouldn't be large windfalls for mining, but that's a good thing.  The windfalls and speculation are what are going to kill it as a currency -- right now it's basically a mix between a commodity and a penny stock, which creates a large barrier to entry from anyone not trying to make a fast buck off it before it collapses. 

I still think it's an interesting project that will be used as a model by future economists, but I'd be very uneasy if I had a large amount of money tied up in them at this point.