Unicode is an all encompassing project, its goal is to make it possible to represent all existing documents as a series of bytes, from ancient hieroglyphics to Japanese txt-speak. Of course, with great power comes great abuse.
It turns out that Unicode includes a complete set of blackletter characters - actually two sets if you count bold. These are supposed to be used to represent mathematical symbols in old documents but nothing is stopping you from using them in Facebook posts to amuse and annoy.
You just need something that will easily convert from normal english text into suitable unicode HTML entities. Something, perhaps, like this:
The output may look like English text, but this isn't the same as just changing the font - the characters are unrelated to the normal ASCII range. You can paste the results into most web pages (Facebook and Google+), but some will strip out anything weird like this. You may have to experiment.
(Before you try it; no, you cannot post blackletter comments on this site. Wordpress doesn't like it.)
Sadly some browsers/devices will not display the glyphs at all, if you only see a row of squares then you are out of luck (iOS devices do not work, neither will Chrome on Windows). Even if it works for you, other people may not be able to see your post.
Another drawback is that the resulting text will not be searchable using normal tools, although this may be a plus in situations you have already thought of.