Kotlin will fall behind on its WASM support unless there is something significant released (has to include significantly improved WASM/JS interop), within the next 6 months that is ready for Kontiners to experiment with, which goes far beyond the WASM sample ( https://github.com/JetBrains/kotlin-native/tree/master/samples/html5Canvas ).

I think what is done currently like Kotlin IR and making Kotlin/JS less tied to JS ecosystem are useful stuff that a potential true WebAssembly support could leverage later

Having additional WASM support delivered in 2020 is too late. Much of the competition (Go, Python, C# etc) already has some significant WASM support in place, even though much of it is experimental. Better to expand WASM support in Kotlin Native (towards the second quarter of 2019), and have it in a experimental state than to create a new Kotlin developer platform, but that is my own opinion.

What Kotlin should target is IMO WASM with GC which is still WIP, I think 2020 is ok for WASM + GC even if you would have prefered it to come earlier