I haven’t tried the script so far, but it looks like a welcome addition!
What I would change is the way information is displayed. In a related (now defunct?) thread I made the following proposal:
If what you have is a number, then one way to do this would be to split the word into kana, and apply the relevant style to the appropriate character. The one thing that might be tricky is that compound kana should be treated as one.
It might be helpful to know that so-called “defficient” mora (long vowels, ん, and the small っ) cannot bear the accent (but are still counted).