Historian says yes!

I would say maybe, but not much is needed. I'm coming at this from the perspective of someone with a degree in physics and PhD in astrophysics, with an interest in various areas of history that rarely have anything to do with the sciences. (Though I am interested in how the early USSR approached the field of statistics, so if anyone has any recommendations there I'd be grateful)

As part of my education we quite frequently did learn about the history of the concepts we were studying. It is useful to learn about older ideas of how things were, why they might have been reasonable to believe at the time, and how they were proven wrong. And also something of the history of the scientists who made various discoveries. I remember one lecturer making sure to point out that the dubious history of the man the Lindemann lecture theatre we were in was named for.

So yes, a history of where the field has come from is useful, and is already present as part of an education in the sciences. But when it comes to actually doing scientific research, no it really isn't needed. When you're trying to figure out a sensible geometry for your simulation or analyze a time-series, the history of the techniques you're using really isn't relevant. The history of physical simulation is likely a fascinating topic for someone to study, but has no bearing on how you go about actually doing it.

And all this talk about the usefulness of history to the sciences has been about the use of the outputs of the academic field. It is useful to know the facts of where things came from, but history as a field is so much more than that. But the practice of doing history isn't really relevant at all to the sciences (except in so far as it is relevant to everything).

The converse really isn't true. The practice of doing history often requires, in part, doing science. Be it dating, statistical analysis of economic data, climate records, or many other things. After all, without that grounding in reality you're just telling stories.