For me, it still feels very early to use Swift for a large-scale app. I've found in my own usage that many of the niceties of the language are offset by painful workarounds to make Objective-C APIs play ball. E.g. computing a random CGFloat within a CGFloat range
On the other hand, it seems obvious that Swift is the future of Cocoa development. And I'm reasonably confident that many of the current issues will be addressed within the next year.
