I've looked into this bug more closely and now it appears that Mike is correct: in fact, this is the same problem as mentioned in the referred post.
I've noticed that the text itself appears exactly the same in both cases (judging by the number of letters, repetitions and double-letters in the same locations), only character sets are different. After some experimenting with Google Translate and character encodings, I think the original text is in Arabic. (Because the last sentence should read "Click Next to continue.")
This means that Thai language was also not detected/supported by the installer. As the result in both cases, installer chose Arabic (perhaps because it is the first language in the list where language names are sorted alphabetically).
It seems like the original Arabic text the installer is trying to display is written in the Windows-1256 (Arabic) code page. Perhaps this would be the default code page on Arabic Windows PCs. However, because in my PC the default is the Windows-1257 (Baltic) code page, I see the gibberish pictured above. And the other customer (Illuminator, in the referred post) has the TIS-620 (Thai) character encoding set as default on their PC. So the characters appear different (Thai script), but in fact the text is completely unreadable because it is originally in Arabic.
To sum up: Arabic appears to be the language always chosen when the installer doesn't support the user's language.