Yes, drifting off the topic
Overall, my recommendation for anyone looking into Kotlin is to not reject Scala, at least without familiarizing a little bit with it. It really depends on the philosophical position, which is really well put in the first few comments here.
Yes, compile times are slower but really fast in-flight compilation in Intelij plugin is available. Java -> Scala calls are not intuitive due to name mangling, but there are ways around it and it is very rarely used anyway. XML literals as many other domain specific languages, which you can create yourself, are optional libraries. It is just more power available when needed. Control flow via exceptions depends on attitude, same as in Java or any other exception capable language, no best practice is encouraging it. Tendency of the community for overly abstract code - yes, seen some libraries like this, which I avoided to use. There is always good and bad code around in any language.
For people coming into Big Data analytics, Scala is inevitable choice due to Apache Spark (yes, there are other language bindings for Spark, but they all way behind in ergonomics and performance).