PDFs have tags for information like Artist, Subject, etc. however MC does not read/write these tags and most PDF authors do not fill these tags with useful information. Therefore you will be relying on folder/file names and anything you enter in the MC library for all information about your PDFs.
My preferred approach is to store PDFs in folders with names that reflect the book topic, and to name the PDF files as [Artist] - [Album] - [Year] where Artist is the author, Album is the title of the book, and Year is the year of publication. After importing a PDF I enter the Artist/Album/Year/Topic tags in MC and then use MC's powerful rename and move feature to put the PDF in its final location. But you need not do this if you prefer to leave the PDFs in their existing location with their existing file names.
To import PDFs that are scattered on your hard drive I would navigate via Drives & Devices to a candidate root folder. Then in the search box enter *.PDF and enable the option "Show Files in Subfolders". You can then select any PDFs you want and right click to import.
One more tip. I have about 11,000 PDFs in my MC library and wanted to search them by content. I did a lot of experimentation with many file indexers and ended up choosing dtSearch. It does a wonderful job of indexing PDF content and has very fast/powerful search capabilities. So I now use MC if I am looking for a book with a certain Author, Title, or general topic. And I use dtSearch if I am searching for information on a specific topic.