I suggested a number of changes to that article and none were made. It's been years since I've looked at that graph but no a Stokes Shift doesn't work backwards. I'll review it and try to recall what Dan did and report back.What's up with the handful of points on the wrong side of the excitation - emission diagonal line?
There's a few points plotted there that have shorter wavelength emission peaks then the light they absorb. Is that accurate? Are they like catching two photons and emitting one?