I could understand that, as we are inclined to hold this for true: "slow -> not smart".

What we do know is "not smart -> slow", because if you are dumb, you will be (infinitely) slow to answer correctly. But note this might indicate that the first proposition is a common logical fallacy where the modus tollens is applied incorrectly.

The slow but brilliant thinker wouldn't perhaps show up for solving a hard and novel problem, as they might have learned they are stupid, and they might still be slogging trough other problem sets. Other excuses are found in https://almossawi.substack.com/p/slow-and-fast-learners-3-qu...

If you want to test pure ability for deep thought, it will be very difficult to control all variables that affect slow people.