So this gets complicated.
Here is a good link explaining some of the differences between rating systems.
PAR Watts is a good way of measuring how much radiative energy is being produced in the relevant wavelengths per plant. You can figure it by taking the raw bulb spectrum and weighting it by the PAR plot I linked above.
The problem with this is that the response from a plant from a high vs. low energy photon (for example, blue vs. red) is often the same. The reason this is so is because a fixed energy input is required for a particular chemical process to occur, and all excess energy is wasted, converted into heat, luminescence, or other modes of energy.
An analogy would be me vs. fdd or browndirtwarrior stepping up to that thing at the carnival where you hit the target with the hammer, and it runs up the pole toward the bell. Say the bell is low enough that I can hit it. Then it doesn't matter if fdd or bdw can hit the target harder - we all make the bell ding - and the fact that they hit it harder might make the bell ring louder or the thing come harder back down the pole. All that matters in the end is that the bell dinged. I am red light, and the other two are, say, blue light.
The reason that this skews PAR watts is that watts is a unit based on energy. So, if we look at the PAR based on quantum efficiency (% absorbed/used photons vs. photons striking the plant), the white line
here, and find that I'm 680nm (red), the others are 440nm (blue), but that we have the same quantum efficiency, what does that mean? It means that we induce the same average number of chemical reactions per photon, or dings per swing in our analogy. When you calculate PAR Watts, however, it tells a different story.
Since light's energy is inversely proportional to wavelength, the blue light has more energy. To be precise, it is in our case 680/440=1.54 x as much energy as the red light.
This means that when computing PAR Watts, the blue photon counts 1.5x as much as the red one, despite the fact that they induce the same reaction. By looking at the PAR plot, however, you can see that most red light is roughly 2x as efficient, so this is partially compensated for.
These differences, however, have led others to compute the meaningful radiation in terms of PFF - photosynthetic photon flux - the number of photons that are meaningful. This eliminates the aforementioned efficiency complications.
Anyway, it's complicated, but PAR Watts are not an altogether bad way to measure the amount of meaningful light. I'll convert those spectra to PAR Watts if I get around to it ... or blazed enough.