From: Reuben Thomas <rrt@sc3d.org>
Date: Tue, 22 Aug 2017 01:51:53 +0100

 ;; Return a string decoded from Nth element of the current dictionary.
 (defun ispell-get-decoded-string (n)
   "Get the decoded string in slot N of the descriptor of the current dict."
   (let* ((slot (or
 		(assoc ispell-current-dictionary ispell-local-dictionary-alist)
 		(assoc ispell-current-dictionary ispell-dictionary-alist)
-		(error "No data for dictionary \"%s\", neither in `ispell-local-dictionary-alist' nor in `ispell-dictionary-alist'"
-		       ispell-current-dictionary)))
-	 (str (nth n slot)))
-    (when (and (> (length str) 0)
-	       (not (multibyte-string-p str)))
-      (setq str (ispell-decode-string str))
-      (or (multibyte-string-p str)
-	  (setq str (string-to-multibyte str))))

Are you sure we don't need to ensure ispell-get-decoded-string always
returns a multibyte string?  What if decode-coding-string returns a
pure ASCII string, which is therefore unibyte?

Rather, Emacs uses a variable-length internal representation of
characters, that stores each character as a sequence of 1 to 5 8-bit
bytes, depending on the magnitude of its codepoint(1).  For example, any
ASCII character takes up only 1 byte, a Latin-1 character takes up 2
bytes, etc.  We call this representation of text “multibyte”.

-- 
https://rrt.sc3d.org