Changing the language specifically requires experience reports. If that's not real world evidence i'm not sure what is.

The problem with the wider go community is they are usually not driven by evidence, but adherence to a philosophy. eg, its not possible to write large programs without generics.