When you use "File->Export text" function on DjVu file with Unicode text, it produces corrupted (all question marks) text file.
To reproduce:
1) Download DjVu file of scanned book (produced by ABBYY FineReader 10)
https://upload.wikimedia.org/wikipedia/commons/2/26/%D5%80%D5%A1%D5%B5%D5%AF%D5%A1%D5%AF%D5%A1%D5%B6_%D5%8D%D5%B8%D5%BE%D5%A5%D5%BF%D5%A1%D5%AF%D5%A1%D5%B6_%D5%80%D5%A1%D5%B6%D6%80%D5%A1%D5%A3%D5%AB%D5%BF%D5%A1%D6%80%D5%A1%D5%B6_%28Soviet_Armenian_Encyclopedia%29_1.djvu
2) Open file with WinDjview (tried v. 2.0.1 and v 2.0.2)
3) File->Export text..
The plain text file produced is not in Unicode encoding, and is unreadable.
If you select piece of text and copy-paste it, it works correctly, so it's not general issue of WinDjview or issue with specific file, guess it has something to do with Text export feature.