Scrape-based search

[wzhouwzhou]

I just wanted to throw out a question to the general community. Similar bigger/bloated packages exist that allow you to search the YouTube API without an API key. This process would involve querying the youtube search results page on your server and parsing the HTML.
Generally, it seems like scraping has remained discouraged, backed up by https://www.youtube.com/robots.txt. If that's the case, those other packages could be violating TOS (Google Terms of Service). Nevertheless, the API and those packages have continued to operate.

---> Should we make an effort to support this behavior of searching without API key? If not, why? If so, how would you want the API to look?