If checking if 21 <= day AND 11 <= month AND 21 <= year is the actual logic they used to check if a date is after Nov 21 2021, I'm not sure they should be writing code that runs on a train.
(Edit: Fixed inequality direction.)
(Edit: Fixed inequality direction.)
Similar stuff is common in PLCs all across the world.
The right test is something like "year < 21 || year == 21 && (month < 11 || month == 11 && (day <= 21))". And yes, probably refactor that logic for clarity, but from this form you can see that it is not equivalent to the test above.
The error is what makes it so that instead of failing "unfixably" once 2021-11-21 happens, it instead simulates failure between nov. 21 and dec. 1, and dec. 1 and january 1, every year starting with 2021.
The condition yields true from 2021 on for:
nov. 21 - nov. 30
dec. 21 - dec. 31