I have taken tests where the questions felt almost repetitive, like I was solving the same type of problem over and over with small variations. At first it felt boring, but later I wondered if that was intentional.
Is repetition used to make sure the test is measuring the same skill consistently? Does it help reduce random errors or lucky guesses? I also wonder if repetition makes it easier to compare performance across people.
This is exactly why subtests exist! They cluster similar items together to measure specific abilities reliably. The repetition also helps establish item difficulty gradients, so they can see if you consistently solve easy ones but struggle with hard ones, or if your performance is all over the place. I agree it can feel tedious though, especially on longer tests where you’re doing 30+ pattern recognition items in a row 
I think the repetition is partly about reliability, making sure the test is measuring consistently rather than being thrown off by one weird item or a lucky guess. When you answer multiple similar questions, it smooths out random errors and gives a more stable picture of your actual ability. It also helps developers confirm that all those items are tapping the same underlying skill, which strengthens the test’s internal consistency.