Disrupting Wall St. implies not just making a "new" UI/UX/Interface for banking clients (like Mint or Square), it implies looking at changing the Banks inherent Business Model. Can Silicon Valley pull it off? Well, from a foreigners perspective (commenting from Lisbon, Portugal) Silicon Valley has a very engineer centric perspective on problem solving, and while engineering inputs may be useful, this is not an engineering problem, it's an economics problem.
Take P2P loans (Zoppa, Prosper, Lending Club) for instance. Why aren't they gaining traction? Because while they try to change the main Business Model for banks, they fail to solve the fundamental problem of Information Assimetry. That's the "reason d'etre" of the Banks. Banks business model is not just "skin you alive in loan fee's", they solve what we economists call "Adverse Selection" problem: how to sort good from bad credit. They are basically information arbitragers. They pool your credit info, compute a score, and sort loan suppliers with loan demand. The cost of doing so is expensive for an individual investor. And there is a problem of "preference revelation", or, in layman's terms, people lie and try to free ride.
That's why banks exist. Is the model ripe for disruption? Yes, it hasn't really changed fundamentally since the last 300 years since the "Venizian Banca" but for that one need to solve the affordable decentralised sorting between creditors and debtors accounting for fraud, incentives to lie and free ride and asymmetrical information.
IMO, Facebook brought a good innovation to the table. And no, I'm not talking about the "like button", social hype (attach social to something and somehow you have an Alchemical transformation of iron to gold): Applied Network Theory.
Social Networking might be a good research direction to solving the P2P information problem solving, by revealing our preferences.
Payment methods have a different problem to it: fraud. You can, and usually do, bleed money on it. Paypal did, and still does. To counter it, you make it more painful to do transactions (that's why Paypal is, sometimes, bloody annoying). Pain acts as a filter to fraud. That's why Banking is so cumbersome. Again, pain as a filter for fraud. Think of it like this: Google could reduce spam by making it painful to search and index (reductio ad absurdum oversimplification). It's a simple "no innovation" solution.
There is a lot of innovation to be made. But it's a bit more complicated. It's not just "make a new cute web 2.0 interface to sort your personal finances" like Mint.
Just my 2 cents.
(PS: pardon the occasional english typing error. Not a native speaker)