I set "levels," and don't include technology beyond the level. Often I only research a word, in an online etymology dictionary, to see when it originated. e.g. "linen" comes from "linum" which is the Roman word for the flax plant, which provides the fibre for "line" (thread). Done, linen predates the time period of my story, thus is allowed, the production of line and fabric is a potential industry.
Basically, although I use modern English words to make it easy for my readers, I don't allow any words that have origins past my cut-off date. If they don't trace back to Old English, or pre-Cutoff, I don't use them, and certainly not any technology they involve.
As for the rest of the research I do, it is often cursory, I look up a few relevant things that prove a point and quit. I intentionally do not take any notes, if I can't remember it for fifteen minutes, it isn't worth remembering. For example, I am not truly interested in all the possible sword types that exist, I don't want to choose one from an array of hundreds, I want one sword that is not too heavy and suitable for a woman to carry. So I am not immersing myself in the history of swords, I am looking for ONE THING to describe, and I know it when I see it.
In fiction, my goal is only to present the sense of the time, I do not feel compelled to be historically accurate. If I give a woman sword X, I don't really care if there is a reader out there that is going to claim I should have given her sword Y. Because sword X is a real sword, I saw it. Maybe that is what her father used to teach her self-defense, and that is her sword of preference. Would a father teach his daughter self-defense? I think it is a pretty radical claim that not ONE father of the time would ever do that! To me it is plausible HER father did that, and -- as always in my fiction -- it is okay to follow exceptional people that do not fit the standard mold, no matter what time period I am writing in.
I research only what I feel is necessary for the page or scene or character. I research with that in mind, and I stop the minute I am convinced I found it. I want to be accurate, but I don't want to write a history lesson or explain why I am accurate. I feel like that kind of exposition stalls the story.
So to answer the question more directly: Nearly everything I research ends up in the book, in the sense that I don't research something unless I really need to write about it, and although I may discover things I didn't know in the process, I don't turn the research into an idle learning process just because it is fun, I stay focused on the goal and the scene, and don't learn any more than is necessary. It is my time to write, not procrastinate by learning a bunch of stuff that will never make it into my story.
Now, that said, I am a discovery writer so I may discard scenes even if they contained a researched item. But at the time I am, say, looking to see medieval footwear, I find that and get out and write about it. I just tried it! It took about one minute to find a history of the shoe from neolithic to modern times, and two more minutes to find medieval shoes specifically, with plenty of ideas of earlier shoes. Roman sandals could be worn today and look normal. Medieval shoes could be basically moccasins with leather or string ties; but you can get fancy with pointy shoes, or even cork soles. Cool. Let's write. That's what my research looks like.