What about implementing text algorithms like prefix search or a suffix tree to mention the simplest ones? Don't you need a string length at various points there?

With UTF-8 you can implement them on top of bytes.