I stumbled upon this issue recently on Github about scraping Twitter data by localisation by combining the rtweet package with the google maps API.
Approximately 1% -