Those simple bar graphs are neat.
Content-wise, I’d suggest to change the survey to differentiate between runtime performance and compiler performance. From skimming the free-form field, I’m not the only one bothered by slow compile times.
Also it might be helpful to state clearly that leaving out answers is ok (e.g. the hiring manager question, or questions about your workplace when you’re not employed).