I don't really find this convincing. It seems like the data comes from a survey conducted by ZeroTurnaround which isn't really scientific evidence. I have a feeling that there is still going to be a significant market for a full featured, free Java IDE for a long time to come, though that's also an unscientific opinion :)
