Are we really doing this for the sake of doing it the Kotlin way? What would happen if you made the implementation just in Kotlin Native and you did JNI bindings for the Java interface? Would getting off the JVM free your performance constraints? From your Graal experiences it sounds like that would be the case

in theory kotlin-native can solve all of this because they do not need to worry about backwards compatibility with type erasure and can do things like specialization without worrying about being compatible. Unfortunately, right now its significantly slow so writing anything numerical directly in k/native to be called from the JVM would make things worse not better.