Doing history is original analysis and narrative. The point here is to meaningfully explain the difference between what a history undergrad is doing and what a professional historian is doing.

Consider an algorithms class in undergrad. You can read about all sorts of algorithms. But learning the Nth algorithm won't transform your work into original algorithms research. You are only consuming information, not producing it. Similarly, just reading other history research can teach you things but isn't what historians are doing. A lot of "history buffs" fall into this category and love to read pop history and consider themselves experts.

Now consider somebody who wants to develop an original algorithm. But they've never learned any analysis methods and they've never critically engaged with the literature. They don't know how to prove an algorithm's correctness or behavior rigorously. There are a ton of these people online. They often gravitate to trying to solve P=NP. This would be comparable to somebody who never learned historiography (the method of doing history) reading primary sources and trying to replicate what historians are doing. Like any field, history has methods. It isn't just ad-hoc decision making from people who happen to have a title next to their name.

In CS this is largely harmless. But for many fields within history, accessing the archive also damages it because people are touching one-of-a-kind objects. So archives are selective in who they choose to allow to access their materials.