I think Ronald is correct.

The priority argument for enter would apply if you called enter twice with two different delays, but they happen to end up pointing to the same moment in time from the scheduler's point of view.

How would the computer know that two calls to enter with the same delay are supposed to point to the same moment in time?

But you are correct, it looks like the example would make more sense if it used enterabs, not enter.

You can test our theory by writing time and delay functions with a course enough resolution that two sequential calls to delay will end up pointing to the time time unit.  (Or we could look at the code :)