Here’s another user who has enjoyed color-coding progress in the past. The parley spaced repetition software used to have color-coding of flashcards, and the same colors were then used as progress bars on flashcard sets (indicating the fraction of words on each of the levels, in ReadLang this could be used in the library). I’d love to see this in the reading interface and I might even be willing to mark up the 2000+ words I already know in the language I’m least fluent in just to get it, but I can’t imagine how much effort it would take to get satisfactory color-coding of the languages in which I’m more fluent.
Words for which a user doesn’t have a flashcard yet could be color-coded based on their frequency in a large corpus (different part of the spectrum than words for which a flashcard exists). This would help the user to decide whether they want to investigate an unknown word or not.
Moreover, at least one other feature of a word is competing for color-coding: there’s a request to color-code noun genders (Suggestion for Improvement: Color-Coding Nouns by Gender).