The primary novel thing Bitcoin provides relative to other systems is inflation resistance.  It's fairly straight forward show that a system denominated in Bitcoin is actually backed by control over the Bitcoin value they claim to control.

(e.g. for example— if you have a distributed payment network based on blinded tokens, you just re-blind and publish all the the tokens, and then publish signmessages with keys in the bitcoin chain for all the coin in the system.  Anyone with a token can verify their their token is in the list... and anyone can sum the tokens and see that there aren't more tokens than coins  ('anyone' in these cases really just being auditing built into the software that runs randomly and publishes alerts if a discrepency is found))

You're adding the centralized requirement there— but it isn't needed, distributed and decentralized systems denominated in Bitcoin can be created— though centralization does allow certain efficiencies. Alternative distributed and decenteralized payment networks systems can scale better than Bitcoin simply because there can be more of them, or they can potentially scale better  because they are less secure (e.g. one using guy fawkes signatures), or because they are less decenteralized. Arguably the latter two are harmful, but better that harm happen on a smaller scale than the whole world and the whole currency.

If the economics of scaling create some centralization I would much rather it be in secondary systems than in Bitcoin itself.  E.g. if to scale to support all soda-pop transactions centralization must be employed to achieve information hiding, I'd much rather have it in secondary system so that _something_ remained decentralized— that there could be choices in centralized systems— and so that people could insist on those systems prove ownership of their backing Bitcoin. If directly scaling bitcoin creates effective centralization there we would have one-centralized-system to rule them all, and proofs of backing would be not worthwhile. Compromising in bitcoin itself would be strictly inferior— better to compartmentalize the compromise if any must exist at all. Nothing can be more decentralized than the ultimate backing asset (e.g. using a bitcoin-like blockchain to move around mtgox-USD would have almost no value over mtgox issuing blinded tokens for mtgox-usd and would be much less scalable), and so if Bitcoin is to back anything it should be as decentralized as possible.

They are— however— the way we signed up for. And so far, no one has yet suggested something else that would obviously work (e.g. the POS stuff is all sadly broken because as Amiller has aptly put "In proof of stake, nothing is actually at stake") much less would be an uncontroversial replacement for what Satoshi created.

Moreover, I'd like to point out that even if a controversial change were technically successful, that success would be potenitally its undoing because it would ultimately erode trust. It would be better for cryptocurrency in general if the succession of incompatible rules were to take place through natural market adoption of alternatives, rather than through a hard-forking change over which there existed any real controversy. ... simply because there is no property right concern when people freely choose to use one system over another, as there may be when a system is, arguably, changed out "from under" people and potentially against their will. (The migration from one cryptocurrency to another is potentially undermining too— 'why adopt foocoin when barcoin will just replace it later?'... but it's potentially less undermining then 'why adopt foocoin when the technorati may change out from under me in the future?'. Considering that Bitcoin's 'competition' is other pure monies backed by the fickle whims of (very powerful) populations and institutions, we must be especially mindful of this.).