I'm not positive on this one, but out seemed the article alluded to the idea that the 100,000hr life was measuring to LED failure and ignoring the possibility o

I'm not positive on this one, but out seemed the article alluded to the idea that the 100,000hr life was measuring to LED failure and ignoring the possibility of driver failure, whereas now there are stricter standards both for measuring the potential of a driver to fail and measuring the lifetime of an LED as the point at which out drops by 30% rather than the point of total failure.