scale of testing != best implementation
The bug recently found underscores that even BTC, with its obvious limitations, isn't flawless or vetted enough. Also, there's the well-known fact, at least in this neck of the woods, that Bitcoin's aim to be an "eletronic cash" hasn't panned out.
If you had one job to do and couldn't do it, would you consider yourself the best?
I tend to fall on the small block (at least for now) side of that argument, and believe bitcoin will find it's way in the end.
Montero does an interesting job of pointing to a lot of possible solutions though...