[bunkr] Fix extracting wmv files

[wlritchi]

The Bunkr extractor already had support for a number of file extensions that Bunkr hosts on a separate CDN. This PR adds another one, .wmv. I also found a few cases of files like foo.MOV which weren't being handled correctly, so this PR also updates the check to be case-insensitive.

[wlritchi]

I'm beginning to wonder if this should operate on a list of exclusions rather than inclusions - it seems to me that there's only so many file extensions that Bunkr could reasonably show a preview for and host directly on the first CDN. We could probably test this by uploading something with a random unknown extension, to see where it ends up by default. Of course, then we'd need to enumerate all of the image extensions that stay on the CDN.

Alternatively, maybe the most robust option is to just HEAD the file, and see if text/html is in the content-type header. If so, it's a download page, so we should fetch it and step through.