> There will be an eventual cap at 21 million bitcoins, so satoshis stash is somewhere between 9% to 4% of all bitcoins.
Btw: As the rules and mechanics of the Bitcoin network are enforced by the majority of clients it would be pretty simple to invalidate these bitcoins if the majority of the Bitcoin software authors would agree to do so. Just ignore any transaction regarding these old addresses.
Of course if a majority agreed to a blacklist, redlist, whitelist or any other exclusionary mechanism, then some coins become less usable than others.
However I think you will find that really nobody wants to participate in such a scheme for a pretty simple reason.
Fungability is an important concept for money in order to work. A landmark case in scottland around 1750 (Crawfurd v. The Royal Bank) recognized that fungability is more important than the individual right in the money. The very same conclusion has been held up in virtually every juristiction imaginable since.
Miners, which very much depend on fungability to run their business, would be dammaging the very foundation on which they are running their business, if they would undermine fungability.
Since you would need the cooperation of the majority of miners, to do something that is not in their self-interest, a serious attack on fungability is not easy to pull off, although it gives credible threat to encourage debate.
Whether this works or not lies largely in the way the Bitcoin software developers pitch their change. For example this cut would be much less directed if you invalidate bitcoins that are older than 3 years. Independent of the Satoshi coins I would favor something like that to clean up lost bitcoin accounts and to shorten the block chain that needs to be cached by all clients. I think Bitcoin is still young enough for such adjustments to work.
The difference between cutting the Satoshi stack and the Crawfurd v. The Royal Bank case IMO is that in the latter invalidating money in active use greatly threatened the viability of the currency whereas in the former the money is just lying around anyway and its existence is a threat to the stability of the currency itself.
For real fungibility with bitcoin we need to add real anonymity, something like zerocoin or some bank like structures that provides huge scale money laundering (in the way this currently works with established money). At the moment bitcoins leave behind such a huge paper trail you can hardly describe any coin as equal to any other, except those freshly minted.
I didn't differentiate between them, because - really - do you think the miners take a close look at the protocol the software they are using actually implements? Much less so the end users.
I see the software engineers of the Bitcoin clients as a largely underestimated force in the network dynamics. Even one subtly placed bug could bring down the whole currency in the long run.
He/She/They should, but having 4% of the total amount of a currency in the hand of one completely unknown person / organization is quite a threat to its stability. Also I would see a reward of one billion dollar as pretty excessive. And that is only the worth at the current market value.
Excessive based on what metric? Inventing the worlds first successful peer-to-peer crypto-currency? Zuck @ Facebook, and other startups (Square etc.) made their founders far richer. Do you realize how hard it is to build trust in a currency even if you are a government?
Strong parallels here to the case of equity in a new business. It wasn't worth $1B when they did the work, and the likelihood that it would ever be worth a significant amount was low enough that the reward would have to be high to be a useful incentive.
In addition, everyone who bought in over the years implicitly accepted this deal. I'm sure if the inventor had claimed 50% of the total currency pool, there would have been less interest.
Btw: As the rules and mechanics of the Bitcoin network are enforced by the majority of clients it would be pretty simple to invalidate these bitcoins if the majority of the Bitcoin software authors would agree to do so. Just ignore any transaction regarding these old addresses.