One thing that nobody's mentioned in the Ethereum camp is how useful good tooling is. If you think back to the past year of Bitcoin history there are a lot of high-profile sites that could have and should have made use of more sophisticated bitcoin escrow transactions. Why didn't they[0]? IMO there isn't going to be much of an abstraction layer over the underlying Bitcoin transaction language and anyone who wants to use these features will have to learn how to write directly in the transaction bytecode[1]. Ethereum will have much better tooling (they're already working hard at it!) and this is how you'll turn Ethereum's competitive feature advantages over Bitcoin into users.

-----

[0] Ok, let's assume good faith on the operator's part.

[1] What I'd like to see is a 'C' for bitcoin transactions instead of the current assembler. I don't think that will happen.