Andrea Gordon has an article in today's Toronto Star about a new kind of software for struggling readers. The software is described as being karaoke-like.
There were apparently three studies by Susan Homan (and others). I can't track down the actual papers, but Homan has links to three "abstracts" on her site. They're rather more elaborate than your standard abstract, especially the first one, which includes the full data set.
The gains do seem to be impressive, the motivational factors are obvious, and the premise makes sense. The words to the songs are already known or mostly known. As they are recalled, the eyes are being presented with the written version. The brain puts the two together and poof! And you there's all that repetition, but it's fun.
I wonder what kind of songs they use and how this could be adapted to an ESL situation.
I've written Homan asking for copies of the original papers and will report back if there's anything to add.