Hello,
I am very familiar with building website scanners.
My proposal is this:
A web crawler (multithreaded windows desktop app) able to
1. find websites from search engines based on keywords;
2. scan lists of web domains, analyse them and collect other domains for future scan, in a circular pattern.
I would go with option number 2.
We will have to work together on the filters, to determine wich websites are eligible for collection.
Cheers.