Paralleisation of `lint_dir`

Linting large projects can take quite some time, however linting each file is an embarassingly parallel task. Maybe there should be an option to use multiple cores using any parallel package?
The limiting factor doesn't seem to be I/O so we should get a lot of speedup from this.

I'd be happy to code up the relevant pieces, preferrably using `future` but open to alternative packages.