Then you start one, because it's foolish to keep developing games outside of a business. If you can't afford to start a business you're way over your head in dealing with EA.
you can do Google Checkout/Amazon Payments.
If you're a business you won't have any trouble dealing with a credit card processor. They'll take a cut, but so does everyone else. It's a matter of making the right choice and spending some time with a spreadsheet.
Then there's the handling of the download - either have to do a login system so people can redownload
You do realize that the guy who wrote minecraft (and he's far from a genius, bless him) actually did all this... Right? Minecraft had sold well over several tens of thousands of copies before he even started pondering about moving his stuff to "the cloud". It was all a single webserver handling it with a credit card processor. Not some sort of magic. Anyone considering making a game larger in size than minecraft has already begun considering the distribution options before slavishly beginning development.
It's the payments and website downloads that are the hardest parts.
If you make a game and you as a company find this "the hardest part", I fear for your business. You don't go to Steam or Origin because you can't handle the traffic or can't delegate payment, you go there because they have a freaking huge userbase and the companies typically have all the credit card numbers of their users neatly stored in a database for impulse purchases. Then after the impulse purchases you wait for the stragglers to buy the game when it's offered in a discount. Go on steam during the weekend, and look at the sales chart and the discounts offered. That's why you use these distribution networks: a large userbase. Despite all that though, there's still plenty of indie devs who succeed in selling their stuff and distributing it without Steam or Origin (or Impulse and its measly marketshare) for that matter.
It's partly why the Apple App Store is very popular
The apple store is an entirely different beast. iPhones and iPads are pretty much walled off from using "typical" software installations. It's the Apple way or no way at all. It's not a matter of choice. And again, you don't start iOS development because there's a neat little platform to distribute stuff over, but because you think your product will have a large userbase and can benefit from the impulse purchases.
These platforms are not going to lower your cost, no matter how pretty they picture it. They have the same costs as you do: (virtual) server infrastructure, maintenance, and credit card processors. They may get a bit of a better rate, but not by much, and don't forget they're out to make a profit too (just like you are) so they are going to be in your pockets for that. So in the end the bill will probably be the same, if not higher because of overhead costs. Oh and there WILL be overhead costs. Don't forget they'll probably make you use their APIs as well, for a seamless integration (friends lists, chat, etc etc). Sure these costs aren't going to be the bulk your costs, but don't forget about that.
You've got to ask yourself when someone offers you this: where is the catch? And it's pretty obvious what the catch is, right? After 90 days those games don't stop existing. Once you're deploying patch 1.12 out on the origin servers they're going to say "Wellp, that's gonna cost you. Remember that contract? Did you read the fine print?". Plan on doing some DLC? Are you sure it's not an Origin exclusive? Bandwidth used by people reinstalling the game after those 90 days? What about your price? Do you get to decide when they offer the inevitable discounts?
I would be very cautious when approached by EA (or anyone for that matter) with such an offer, because you can be sure there's no such thing as a free lunch, and you can be damned sure that EA of all companies isn't doing this out of the kindness of their heart either. They're out to maximize THEIR profits, not yours.