So the first 4 columns are based on the manufacturer specifications. Two of them did not rate the lumens at all, so they are blank. The next rows are recorded data.
The exposure column indicates how many seconds of exposure was needed to get a normal photo exposure from my dSLR camera with a fixed aperture in a light-sealed closet lit only by the tested light bulb. Thus 10 means 10 second exposure, 2.5 is a 2.5 second exposure, and so on. The shorter the exposure, the more light the bulb produced.
The next 3 columns were a result of plugging in to a standard lamp plugged in to a Kill-A-Watt device and allowed to warm up for at least 2 minutes before taking readings. Oh yeah, the exposure times were after a 2-minute warm-up too.
The findings of interest to me are as follows:
1) All the CFL's except for 1 used less wattage than they are rated for, which is kind of good for your electric bill, since that's based on wattage. A 60-watt incandescent bulb drew exactly 60 watts, but most all the CFL's drew between 1 and 3 watts less than they were rated for. The 65-watt megalight drew a full 12 watts less than it was rated for.
2) Regardless of the lumens output, CFL's do not make as much "real" light as incandescent bulbs. I have observed this just by the naked eye that a 14-watt CFL rated at 850 lumens still didn't appear as bright as a 60-watt standard light bulb rated at a mere 630 lumens. Sure enough, the 60-watt bulb required only a 2 second exposure and the 850 lumen CFL needed a 2.5 second exposure to match it, indicating less measured light. They traditionally say that CFL's produce 4x as much light per watt as incandecents, but I've believed it is more like 3x as much light. It took a 20w rated (19w observed) CFL to get the same exposure as a 60-watt incandescent, so I think the 3x generalization is more realistic.
3) The power factors on most CFL's is horrible. They ranged from a worst of 0.48 for one of the 4-watt bulbs to no better than 0.62 on the best of the normal CFL's. While only wattage goes against your power bill, your power company has to produce VA worth of power, not watts, so the Eiko 4U20 for example, although it draws just 19 watts, it drew 30 VA, which means it only represents a 50% power savings at the generating station, not the 67% savings from your power bill.
4) Two of the CFL's had almost perfect power factor correction, those two being the two highest wattage rated bulbs, rated at 27 and 65 watts respectively, although consuming 24 and 53 watts measured. Both had almost perfect PFC, meaning the 27 watt rated CFL pulling 24 VA actually requires about the same amount of power to be generated at the generation station as the 14-watt rated CFL pulling 23VA. From an environmentalist standpoint, a single 27 watter would definitely be a smarter choice than two 14 watters, for example.
In conclusion, I was glad to see that there's at least a couple CFL's already on the market with fantastic PFC or whatever it takes to show a 0.99 power factor. I expect that as PF is regulated more closely, we'll see more and more applicances and lighting with this trait. At least we know it's already possible with currently marketed bulbs.
