Thanks for forwarding this, Jack. It is very interesting! Though they (and you) are careful to note that "This survey should not be viewed as a systematically representative sample of the astronomy community", primarily because 63% of the respondents were either grad students or post-docs. I'm glad to see them acknowledge that they aren't trained in (and haven't attempted) correcting for the significant selection bias.
Not surprising to me at all that Python is the preferred language for the younger crowd, and thus eventually (even if just due to the passage of time) for the whole field.