Is it feasible to create a word board that illuminates in time to a song?

I've seen a lot of projects like this one:

My goal is to create a similar board with the words in a song, then illuminate the lyrics in time with the song as it plays. I can fit all the words on a grid of letters 20x20. I'm hoping to use an Arduino or something similar to illuminate the words and play the song through a 3.5mm jack. No more than 20 LED's would need to be lit at any given time. I haven't been able to find any similar projects that people have done. I'm concerned about playing the song and lighting the lights at the specific points they need to be lit. Is this feasible?

The word board is relatively easy.
How will you determine the proper time to illuminate a word?

I was planning to play through the audio and manually determine a time and duration for each word. Yes, I know that'll be time-consuming. There are just under four hundred words, and just under a hundred unique words, some of which fit into others (e.g. 'the' fits into 'there')