SpringSource acquiring G2One and Groovy and Grails with it is a strong indicator to me that there's a need for other languages (in this case a scripting language) on the JVM. SpringSource has certainly been an indicator of what tools will be successful in the Java ecosystem for some years now.
Scala has features that Java will probably never have, and even features like closures are a long way off for Java. That's true of many of the new JVM languages. As developers are exposed to these features, it's hard to go back to Java.
You can point to major companies like Twitter, FourSquare, Sony Entertainment, etc who are using Scala in production, but to me the more telling piece is that developers are excited about these new languages and see value in the features they offer beyond Java.
I think Android may help out one or more of the new JVM languages a lot. I know a lot of people are writing Scala Android apps.
Java's obviously not going anywhere any time soon because there's a huge investment in it. So the JVM is around for a long time. But as a a language Java has started to lag behind. To me it seems very likely that new languages that offer concrete advantages over Java, but can be deployed to the same environment you already support will thrive in the developer community.