When multiple people taste the same coffee under the same environment and cupping protocol, how closely do their notes typically match each other?

And how consistent are their tasting notes if repeated (blindly of course) on a different day or later in the day?

The tasters must have sensory evaluation training to a consistent reference. Probably the most well-known system in coffee is Q-Grader certification. Here are a couple of papers that I found on researchgate that look at the consistency of Q-Graders: (It is free to join researchgate, but these links should work for everyone.)

The consistency in the sensory analysis of coffees using Q-graders

Very beyond subjectivity: The limit of accuracy of Q-Graders

Abstract only: Adherence and concordance among Q-Graders in the sensory analysis of coffees

... and a bonus link:

Does Coffee Have Terroir and How Should It Be Assessed