Seems like more people are catching on to the fact that lumens aren't really all that matters these days. But the scary thing is that even most people who know that don't know just HOW wrong lumens are as a rating for grow lights. It's not just a little wrong, it's actually harmfully misleading. Luminous flux - Wikipedia, the free encyclopedia (the black line on the graph is used to rate a light's luminous flux) As you can see, it is an incredibly skewed representation of the actual energy output of the light. Wavelengths in the deep red and high blue spectrums are rated at nearly 0, regardless of energy intensity. And of course those ranges are the MOST photosynthetically-active. But perhaps some of you have seen this graph for PAR: http://www.efn.org/~k_mccree/Biographical/BioProf/BioProfImages/FigPAR.jpg It's comically wrong. And surprise surprise, it's overlaid on the spectrum graph of every HID bulb I see. Makes HPS look great with the fake peak at yellow. It also shows green light as MORE active than blue light, which is pretty comical. Green light is of course non-active in plants to any real degree since it's reflected by chlorophyll. Lumens overrates HPS and MH bulbs while underrating lights with better spectrums. HPS gets overrated the most by the lumen system since sodium throws out its peak at yellow. I'm sure that's also why people don't use red-spectrum HPS en-masse. It won't be long before people figure out that targeted fluorescents (most major fluorescent manufacturers make pure red/blue T5s) and LEDs vastly outperform HID systems at equivalent wattages. I'm happy to see many other growers on this forum are discovering this as well. I was repeatedly told that fluorescents would not give good bud density, there wouldn't be any canopy penetration, etc. I did a small grow with 3 125W CFLs and everything claimed was completely wrong. I'm guessing people who have had poor results were hanging the lights feet away from the plants, and/or trying to use 125W of fluorescents to replace 400W of HPS and being surprised when it doesn't work.