I think the future analysis of the blockchain (all Bitcoin transactions are public record) will be a ripe area for economic research. Aside from the technical stuff, watching the growth and mistakes of the economic system is fascinating - like seeing 2000 years of banking/currency evolution packed into a few years. There is also the question of intrinsic worth and how to value a bitcoin - most online debates descend into chaos but the theoretical basis for projecting valuation is a real mind bender. There is no precedent.
I'm personally interested in what, if any, nation state involvment there has been with Bitcoin. Regardless of Bitcoin thriving or dying, the system has unequivocally demonstrated that cryptographic systems relying on decentralised trust can work. The theoretical underpinnings of Bitcoin solve the Byzantine's Generals' Problem. And it doesnt just apply to money, it can apply to anything. There is already a decentralised DNS (NameCoin) based on the same protocol.
I would have thought this type of system would be of keen interest to many nation states, as it could potentially undermine their own authority or be used as a weapon against other nation states.
So yeah, a book would be interesting :) I strongly believe the decentralisation technology underpinning Bitcoin will be a notable part of history.
*In 2011 a 'msytery miner' nearly doubled the total network's computing power for a short period only to disappear within a couple of days. The theories are this miner was a clever botnet utilising host GPUs,a supercomputer, or someone testing very early stage ASIC hardware, despite none of the known ASIC producers having working hardware at that point in time. http://bitcoin.atspace.com/mysteryminer.html