Scala is the first language I've seen where the OO and functional features are on an even plane and even play nicely together, and where the community is accepting of both styles of programming.
Clojure has a couple major things going for it. One is that, to me and at least a few other people, it feels like a well-designed tool. I constantly get the sense, when programming in Clojure that Rich Hickey understands what I want to do, and designed a tool to make it easier.
The other big advantage is its focus on concurrency and parallelism[0]. Clojure is built around constructs designed to make concurrent and parallel programming easy. All the data structures are immutable, but relatively efficient. The built-in constructs for coordinating state changes are very well thought out. These features could be added as libraries to most languages, but having them built means libraries will use them too, and the syntax is nicer.
[0] They're not the same thing: http://ghcmutterings.wordpress.com/2009/10/06/parallelism-co...
I think the recent popularity of Scala and Clojure, vs other languages that mix the OO and functional paradigms, really come from one thing and one thing only: Scala and Clojure run on the JVM. Everything else is really secondary.
Well, there's that and some kind of successful marketing effort going on. Unfortunately, it's not easy to study exactly why "buzz" builds up around certain languages. Clearly Java and C# had whole corporations' marketing departments behind them, so that probably didn't hurt. Ruby and Python made "Perl sucks so you should program in us instead" in to a successful mantra (you know what they say about repeating a lie often enough..).
Clojure and Scala don't really have the same things going for them, but they are buzzword compliant, mixing OO, functional programming, and the JVM together. Programmers hear that and it sounds sexy, so they jump on board.
Now the question becomes more of whether that initial buzz can be converted in to long-term success for the language, which depends less on language features than on the language's community and on further marketing (which, of course, may depend on having "killer apps" and what companies, projects, and personalities they can get on board).