Our data is already publish via torrent, see https://planet.openstreetmap.org
Our data, including minutely updates is available via public S3 buckets (EU & US) supported by the AWS Open Data Sponsorship Program.
As expected. Now for the hard part: how to redirect scrapers to use torrents, and keep them uploading data? Should we write some "data scraping tutorials" specifically targeting OSM and describe using torrents as the best method?