I think your approach is slightly wrong. First we formulate a thesis statement. Something we are trying to prove or disprove. Based upon this thesis statement we then branch out looking for primary and secondary sources that will be useful in that thesis - and then we pour over it with a fine tooth comb looking for information that can be used in our arguments. So it isn't so much that we skim documents but rather we look for documents that have some relevancy and pertinence to our topic, and then thoroughly read it.

Scrotum_frog is correct for academic historians- if you are going to publish on a specific question in your field you have to be familiar with both current theory in that field and do research in the sources needed to answer the question. But though there are historians who will move from place to place ( H.R Trevor-Roper wandered into all sorts of places), there have also been ones who stayed pretty much in one place and acquired a very good familiarity with the sources. For some areas like Medieval history that can be necessary: the sources can be few, difficult to understand or translate, and can have had extensive fierce debates over them, and it can therefore be hard to jump into them unless you master a lot of information. If I wanted to go from 18th c. technology ( where I am comfortable) to 16th c. technology I would only have some work to do, but if I were to stray into 6th c. England I would need an immense ( maybe now impossible) amount of preparation.