I have very little in the way of paper documents, except for what was passed down to me in the family, so for the most part, this reply is talking about the folder arrangement for the digital folders on my computer, both for the scans we made of our own family papers, and the images downloaded from websites (what Ancestry calls "citation media"). But the same principles can be applied to a paper system.
I keep my document image copies in archival arrangement, sorted by where I got the images from.
I was prompted to do this early on, when I discovered two brothers living next door to each other on the census -- not from reading the census page, but by downloading the digital image, and discovering that I had downloaded it already. That taught me that 1) I wasn't spending enough time to analyse the whole image, and 2) if I saved a copy of the image for every person in the census household, I would quickly be up to my ears in duplicate images.
I know it is popular to keep everything in folders associated with the particular people. I do that also -- for my research notes, not for the image copies.
The reason I do it this way is to provide separation between my own work product, and the material which comes from another entity. Suppose I wanted to publish a book, and needed to ask for permission to reproduce an image. If I rename that image (as so many people do) to indicate the name of the person in the image, and store it in that person's folder, I now have no idea where I found the image, and I have to go out and look for it all over again so I can know who to ask for permission.
I have heard from some professional genealogists (if I recall correctly, it was a post on the RootsWeb mailing list TRANSITIONAL-GENEALOGISTS-FORUM-L, but it might have been in the Q/A session of a webinar) that in some instances they were not allowed to provide their clients with the documents they had looked at for the client's research, and could only provide the information for the client to access them with their own personal subscription!
I use several methods to keep track for all the records I have about a single individual. For the images where I'm sure I have a good identification, I refer to Family Historian to see what documents I have gathered about a person. Also, when I am working on a problem, I create a Genealogy Source Checklist like
this one demoed by Crista Cowan on her Ancestry Desktop Education playlist on YouTube, and place it in that person's research notes folder.
Before I'm sure where a particular source belongs, I make notes in Scrivener. And as part of my review cycle, I plan to inventory all my media and enter it into the source-centered software
Clooz.
I also have
Evidentia, which walks the user through the process of writing a proof statement about specific research questions. It produces reports too, and those can be stored in a person's research folder with my work notes.
If you want a paper solution, instead of relying on the computer to cross-reference your records, then what about making a Source List for all the sources you've found so far? Assign a number to every document that comes in, and list it on that person's Source List. You'll be able to see at a glance what records you have for an individual.
The FamilySearch Research Wiki article on
Research Logs has an example of a simple research log with a space for a document number -- the blank form is available for download
here. If that log doesn't suit you, there are many others available for download or purchase. The Wiki article includes links to videos in FamilySearch's learning center where G. David Dilts demonstrates how he keeps his own research log. Do what works best for you.