I agree with those saying that the test is pseudoscience, but what makes it dangerous is not how it is measured, but how it is applied to somehow categorise workers/students. I've been in a lot of workplaces around a lot of people (and I'd say I've had less jobs for someone my age than most), and 'intrinsic personality' is one of the least useful ways of determining how well somebody is going to do their job. Almost every other factor (ranging from external factors like pay, how interesting or relevant the work is, how supportive the team they're on is, through to internal but fluctuating factors like how much sleep they've had, and what they've eaten) is vastly more important than whether they are 'introverts' or 'extroverts'.
I think of the test results as being slightly more useful than astrology, but not by very much.