The author basically argues that things should be wonderful and easy and the service providers should handle it all (ignoring the difficulties of such a provision):
I should use a plastic card, cheap, easily replaceable, low cost (free in most cases), and my card account should buffer my purchase as long as the combined total of credit/cash in my accounts is greater than or equal to my purchase. I should then be able to place either the entirety of my purchase or parts of my purchase in separate accounts that provide different benefits. (Think, business, flight mileage, cash back rewards, etc) The service should often be smart enough to learn where I move my purchase and do so automatically if I so choose.
and
I believe we should be moving away from cards altogether. I think google wallet, square, and other RFID/NFC technologies are thinking about the future, where our devices are consolidated and integrated with the world around us. Bitcoin, a virtual currency, proves the success of something that exists purely in the aether, and keeping a Bitcoin wallet on your phone is easy.
How is going from 'wallet filled with cards' to 'one card' not a step in the right direction to 'no cards'? Maybe the technology or industry isn't moving as fast as the author would like, but it's entirely hyperbolic to say that Coin's trajectory runs opposite from the author's ideal vision.
Coin's goal is to abstract the ownership of a credit card away from the plastic itself. If you want RFID/NFC to succeed -- as the author does, as I do, and as I'm sure many other people do, then that's a good thing.
(The preorder legality side of things isn't my domain, so I can't really comment on that.)