Super interesting stuff. I am currently exploring a solution where I infer the word usage using the wikidictionary dump. Other solutions would be to make my own (or let user submitted a daily grid). I also want to try using some available word frequencies, but I need a solution that supports French, Spanish, German, etc.
wordfreq, if it's a thing you can use, definitely has information for French, Spanish, and German. (If you don't use Python, the data files are available in MessagePack format at https://github.com/rspeer/wordfreq/tree/master/wordfreq/data, which ought to be readable in the programming language of your choice.)
I know Elia, who wrote wordfreq, used wiktionary dumps for other projects, and I know there's useful data in there (I think among other things she was filtering out offensive words?), but I don't know much about it. Best of luck!