Never did figure out if it was stupidity
or malice driving that.
Personally I think it's neither; it's just $$$$$.
They could invest in a robust system to accurately identify what they chose not to allow to access the service. Or, they can choose to run with a 'close enough' system with some legitimate users caught in the middle.
They've most likely done the math and decided that the revenue lost from people getting caught up in inaccurate blocking is small enough that the investment in a more accurate method isn't worth it. This is unfortunately the more common decision in this age of worship at the Altar of Maximum Shareholder Value.