@Melbustus:
You're talking about an improvement on gold, not fiat. Controlling the money supply is a useful property of fiat, a tool not available with Bitcoin.
TL;DR: Deflation is not "a good thing." See
http://en.wikipedia.org/wiki/DeflationI'm talking about an improvement on both. I understand full well that the branch of monetary economics most popular with governments over the past 100yrs or so has considered an elastic money supply to be beneficial. I also understand why in some depth (econ degree (which should make me naturally hate bitcoin, right?)). The math is indeed enticing, and does work in constrained environments. If central banks (ie, the handful of people running them) always operated with absolute perfection, in near-ideal academic circumstances, there'd be a strong argument for that being the better path (and there would still be strong counterarguments).
But there are baser problems. CBs don't react fast enough, don't have perfect information, are influenced by politics whether they're officially "independent" or not, often have multiple competing mandates, and operate in a literally chaotic mathematical system *as if* the system they're actually manipulating is fully describable by a simple linear relationship. So, arguably, the effect is that while they may smooth medium-term business cycles in non-outlier cases, they create far bigger instability, which, while appearing less often than the natural ups and downs of the business cycle (also debatable), have potentially high-impact outcomes that could lead to longer-term systemic problems, and in the worst case, geopolitical issues; ie, war.
The knee-jerk "deflation is bad crowd" (thanks for the wikipedia link, how helpful) often doesn't realize that their arguments often hinge on an underlying assumption that the *overall* *longrun* money supply dynamics have an inflationary bias, and are not known to, or predictable by, all economic participants. Thus, when some shock *does* create a deflationary moment, it gets exacerbated by economic actors who freak out because they (correctly) realize that they should take this brief opportunity to hoard. IFF the money supply dynamics were *perfectly* known to all economic participants, inflationary, deflationary, whatever - just known - everyone would be able to allocate against that backdrop, and the "what if" feedback loop wouldn't form with such ferocity. It's the same equation, where you put the elasticity doesn't matter in the long-run. Thus, the other considerations above should be given more weight.
At this point, Bitcoin's inflation rate is way higher than USD--that's what mined coin is. Bitcoin supply will continue to inflate throughout our lifetime.
As was pointed out by Satoshi, the price of Bitcoin tends to the cost of mining it (and vice versa).
Mostly the vice-versa part, but moving on...
In other words, the price of producing one bitcoin tends to the cost of producing it. This means that if Bitcoin ever becomes substantially more valuable, the value dumped into mining it will proportionally increase.
Do you think that ~10% of the world's wealth is going into maintaining today's fiat "ledger" every year?
The amount spent on mining will ultimately be slightly less than the amount of transaction fees generated by the network. If bitcoin becomes a high-volume system (see Gavin's latest scalability roadmap), transaction fees can stay low relative to transaction size and still keep the network secure. The cost spent on mining would be completely acceptable and rational.