I think people are annoyed about the idea of having to know R, Python and Julia. So we stick to our guns with Python even if Julia is possibly better. Heck, if I'm ever learning another language, it's Janet then Rust. Or Type/Javascript if life ever throws me thataway. To paraphrase: stop trying to make Julia happen.

Also the Julia community is lovely, but R folks are so smug about tiny things like (the admittedly very fine for its applications) ggplot2. Also: have you heard SAS or Stata users telling you they can run neural networks using their built-in matrix languages? Yes we can! Yes we can!