One reason would be if you have coded a robot to scrape http(s) content of other sites, its easier to configure wget/curl/own_code to use your proxy (or even a list of own proxies) than seting up vpn,
see my post above in this thread about setting up tinyproxy
can also use haproxy if its only one specific site you are scraping
