I'm playing devils advocate here but I think you guys underestimate the banking business, particularly in today's world.
Calling the banks stodgy or tech-averse is superficial IMHO. Banks are not tech averse, they are incredibly financially disciplined (it's their raison d'etre!). These are not VC shops looking to cash in by inventing the next best tech.
Additionally you are ignoring the regulatory component. The FRB will be so far up a new bank's colon that it's too painful to think of. The large banks took 10 years to comply with FRB/OCC matters. And they are still not done.
Citi, Wells, JPM all have over 200,000 employees, each. Google and apple have half that. Simply to bring on the kind of manpower needed will be a billion dollar endeavor.
It is not just 'lets build some credit models and start slingin' cards". On the models side, these banks have 2, 3, 4 thousand models each. Who is going to build these? Who is going to validate them? How long is that going to take? It is not feasible. And the people building these models - they are not cheap. We billed out at 600, 700 an hour for regulatory models, not even valuation models which are much more important. And then you need a validation group which again, is incredibly not cheap.
Then you need to integrate into the markets. On the consumer side, now you need instantaneous scoring services at a massive scale and you need retail partners to integrate it. Retail partners who are already being serviced by these large banks and at a lower cost than you can provide. And these banks and policy teams who know these business better than you (and they) do. This is not just incredibly expensive but a hell of an endeavor to start from scratch. Systems migrations take two years - and that's a migration.
On the institutional side, it's even more opaque. First you have no idea what these product which are being traded. Show me anyone at Apple or Google who can explain why Kirk's spread option model is conceptually unsound but under what circumstances it is still acceptable to use. Nobody. Now tell me who is going to figure that out and then design a systems application to price certain products with certain models under specific circumstances. Of course you can use vendors to tap into the market in this way but the regulators will destroy you. And they're expensive as hell and there's a reason all the banks have migrated to in-house solutions. So unless you want to lose money on every trade for 4-5 years until you can build your own system to migrate over from a vendor platform, you're out of luck.
This article:
https://seekingalpha.com/article/2561895-apple-pay-has-long-term-implications-for-visa-mastercard-and-retail-bankspostulates that Apple is entering the retail payments sphere as a means to enter the financial industry at a whole.
Well look at the payments - has Apple or Stripe built their own systems? No, they are playing on top of the established rails. Maybe this will change but it is difficult to see why. To build out such a system is incredibly expensive and the payoff is very uncertain. Expected ROI today is almost certainly very negative. And V/MC and the banks are expected to sit tight while this happens? I think not. Even if Apple or Google does go down this road they are opening themselves up to so many costs it will be absolutely brutal and the banks will slaughter them on the institutional side.
Anyways take it with a grain of salt because predicting the future is a foggy endeavor but if I had to wager I would say the odds are with the status quo.