Note that while the stored URLs are standardised to use
http:// we actually use protocol independent links—e.g., for the linked license URL we link to
//creativecommons.org/licenses/by-nc-sa/3.0/ (which equals
https://creativecommons.org/licenses/by-nc-sa/3.0/ for anyone seeing it on https://musicbrainz.org). We standardise a lot of the URL formats so that it's easier to look them up from programs or scripts. One of the things we usually normalise is the protocol, to
http: or to
https: depending on a number of factors. I think Creative Commons use
http://… themselves for their canonical URLs, which is why we store them that way (for now).
You can look up URLs and then query those URLs individually. There's a community tool which allows something like this:
However, I think MusicBrainz's "internal" search also allows for URL searching now, even if it isn't documented. I'm not 100% on the details though (since it isn't documented). Once you have found the URL(s) somehow, you can use
/ws/2/url browse requests to find entities linked to the license.