Comparing download times of base R and the curl package:
url <- 'https://jeroen.github.io/data/nycflights13.json'
tmp <- tempfile()
system.time(download.file(url, tmp, quiet = F)
system.time(curl::curl_download(url, tmp, quiet = F))
Because this is an uncompressed file, the gzipped download is about 7x faster.
For the solution see https://curl.se/libcurl/c/CURLOPT_ACCEPT_ENCODING.html :
> To aid applications not having to bother about what specific algorithms
this particular libcurl build supports, libcurl allows a zero-length string to
be set ("") to ask for an Accept-Encoding: header to be used that
contains all built-in supported encodings.