we should not make background requests for remote sites

[bridiver]

For some media pubs we make extra requests to urls the user has not visited to extract things like channel id and favicon. We should try to find ways to get this information using existing requests. At the very least we should minimize the number or find other ways to get the data we need.