This MIME type at the server level seems to worth what it worths. I used a test server. I removed the .pdf MIME type entry. I refreshed my page, it still works. Then, I added it back and tried to mix it by indicating application/octet-stream and it still worked. Then, I changed it to images/jpg and it still worked. So, there or not or no matter the type indicated, this does not have any impact on how the browser interprets such file.
I see all the MIME types defined at the server levels. Then, I went at the Web site level and try to add a bypass at the header level by doing the same test. Still the same result I obtained. I am really questioning myself as to know if this really serves for something.