It seems that the server is sending a UTF-8 encoded document, but the http header declares ISO-8859-1, which is usually interpreted and displayed as windows-1252 by browsers.
If the document has the "right single quotation mark" apostrophe (
U+2019) as in "AKULA's" (which I won't try to actually enter properly now), it has bytes 0xE2 0x80 0x99. If the document was declared as UTF-8, you'd see the correct apostrophe. When it isn't, you get three one-byte windows-1252 characters which happen to be a-circumflex, euro and TM. (The latter two don't exist in ISO-8859-1 itself.)
I'm checking this on Firefox which has removed the option to change the display encoding manually. Nowadays it only has a "View -> Repair Text Encoding" option, which
may detect the document as UTF-8 and switch the display encoding to that,
if there are clearly invalid characters visible. The change should show in the "Tools -> Page Info", Text Encoding field. But if you force the encoding by other means, such as editing the header to UTF-8 or downloading the page and opening it in some editor which uses or allows UTF-8, the text looks fine.
Unfortunately, I don't know what happened and when and why, and what it would take to fix it. I'm not a web server or SMF admin enough.

(Posting this anyway despite the new message. I spent many minutes on this and even my coffee got cold.

)