Ripple is very difficult to understand and therefore is difficult to argument against it.
But what I understood it is centralized and not open source.
Another problem is that it needs gateways like exchanges by bitcoin and actually there only one gateway.
Without gateways to fiat you cannot transfer fiat money just Ripple.
So the problem which exists by bitcoin with exchanges by Ripple is even worse at the moment it is just hidden by the complexity of the system which is difficult to understand.
This is pretty much all true. But none of these things are fundamental about Ripple. It's just the difference between where we are and where we are going. You basically just summarized our priority list right now.
The XRP currency is somewhat similar to bitcoin, a fixed supply of coins that are owned by the company. The same deflationary characteristics only with a central issuer that hopes to earn money from seigniorage. So instead of individual speculators riding the deflation bubble, there is a single company dedicated to this, conjuring money out of thin air and playing the role of the central bank. I have to say the scheme is pretty sleazy and smart at the same time, but hardly revolutionary. If Ripple is successful there will be tremendous pressure from regulators to break or cripple it's anonymity and the company will have no choice but to comply. It's utterly irrelevant if the source is open, since the company owns the "central bank" keys, and there will be no reason to move to a different network with other keys where none of your money are valid anymore.
The one part I find exceptionally interesting is how the network achieves consensus without burning resources, using a darknet approach: nodes trust each other on an individual basis and keep their trust network secret. It's great that research and practical experimentations goes into this area, and I absolutely love the continuous ledger closing. I am a bit skeptical that the particular approach can achieve long term consensus in the face of network splits and deliberate attack, and I have my bets on prof of stake as the correct approach to this problem. But still, a practical way to run a distributed cryptocurrency without mining is a great experiment, proving there's no need for waste. If the scheme withstands scrutiny, it could form the basis of a truly open currency, where the minted money can go to the keys of a set of predefined charities, instead of private investors. (anyone interested can send my a message to brainstorm about this)
As for the credit network functionality, my primary gripe with "ripple type" systems is that they break the fungibility of money. I no longer have X amount of coins, I have a portfolio of credit lines of varying solvency. So instead of just trusting the bitcoin network (developers, miners etc.), I have to trust the ripple implementation (Opencoin, etc.) and on top of that I need to trust individual credit issuers when accepting credit from them and that their credit lines will remain solvent for some time until I will spend them. When you say "X dollars in ripple", the "dollar" there just fulfils the "unit of account" function of money. It's not a store of value and it's not a (fungible) means of exchange, in other words I can't reliably express my time and purchase preferences as I can with a dollar or a bitcoin.
Once people understand how ripple works (probably never for most people) the natural response is to settle al credits immediately to the entity trusted by most people, so the credit market will tend to form a cartel or a complete monopoly. This is not unlike the current banking landscape, only without state regulation. When a trusted entity will fail - and it has allot of incentives to do just that - it will wreck havoc. Maybe there's value in the ripple credit network, but ripple credit is not "money" in the common sense of the word: debt issued by an indisputable single entity with the sole purpose of serving as an exchange medium and with zero incentive for manipulation.