En version courte (l'article est long, mais illustré de quelques exemples qui valent le coup d'oeil) :

"Yes, a Bitcoin fits many of the theoretical definitions of a currency. (...) It needs to be a portable, divisible, easily recognized form of exchange that is universally accepted by merchants and service providers.
There are several other key facets of a currency that need to be taken into account in the modern day. First off, we have the matter of inflation. If you earn $100 and put it in the bank, then inflation acts as a sort of virtual tax on your money. It’s worth less and less each day. But, once you get past that basic description of inflation, you realize that inflation is actually a good thing. Because the opposite of inflation is something far more terrifying and sinister: deflation.
As it turns out, Bitcoin guarantees that deflation must occur. In the year 2140, the money supply will freeze at 21 million Bitcoins in existence. Decades up to that point, the supply of Bitcoins will basically stagnate.
But does this mean that Bitcoins will necessarily experience deflation? Well, yes it does. If people actually use it, then it will deflate.
Because if you are not increasing the money supply of a currency, then any increase in productivity will lead to deflation. Simply stated, there is an increasing amount of goods and services chasing a flat supply of money… this means that each coin is worth more than before."

Great, now engineers think that they are economists too (10/04/2013)

Dans le même ton que l'article ci-dessus, on peut lire également :

Pour la contre-argumentation, voir par exemple :

Un très long article de Felix Salmon, assez neutre de ton mais qui expose clairement les risques :

Côté français, Georges Ugeux est plus vindicatif, et parle d'ailleurs davantage de fraude que du risque de déflation :

Et pour rappel, la meilleure (amha) introduction :