Note there is an error in the paper going from equation (60) to (63): inequality (63) should read 10^-8 Hz < mu < 10^2 Hz. This is fairly important from an observational standpoint!