Judging unemployment via the unemployment rate is a bit like judging your website's popularity via Alexa. The question becomes: is lousy data better or worse than no data?
It's not no data. There are lots of data points to use, it's the choice of exactly which data points to use that is in question. The same debate rages over inflation as well.
The issue is who is and who isn't included in the definition of unemployed. The truth is, if you don't have a job and you need one, but you've given up your search, you're still unemployed.
Further how do we count people who are underemployed? For instance, software developers who take fry-cook jobs because a crappy paycheck beats no paycheck. Self-employed is tricky too, because is slow business unemployed? Sure each of these edge cases is just that, but there are enough to add up to a noticable amount.
The issue is who is and who isn't included in the definition of unemployed. The truth is, if you don't have a job and you need one, but you've given up your search, you're still unemployed.