Maybe measuring semantic density would make more sense than literal characters/words. Produce two matching texts in Japanese and English that multiple experts i

Maybe measuring semantic density would make more sense than literal characters/words. Produce two matching texts in Japanese and English that multiple experts in both languages can agree are comparable in both meaning and relative verbosity, then divide the total number of English words by the Japanese typer's time, and you'll have a rough WPM score that can be used for more apples-to-apples comparisons.