A simple way of saying this - companies are optimizing for filtering out bad candidates, at the expense of sometimes filtering out good candidates.
Because the cost of hiring the wrong person is a LOT higher than missing out on the right one.
Because the cost of hiring the wrong person is a LOT higher than missing out on the right one.
They might not be filtering out the bad at the expense of the good. But filtering out some of the good in the name of saving the money it would cost to develop / administer more general assessments.
That's the p(interview_capable) piece whereas the trade-off you mention is the conditional probability (also important!).
Why can't anyone come up with a good solution for this? A "we'll hire you for a month and see how it goes" kind of deal?