I very often solve this situation. I find an interesting paper on the Internet and importing them into the Zotero. In next steps, I read references in the downloaded document an if I find some interesting or important references, I import them too. Unfortunately, in a lot of cases, I find these documents in the duplicate items because I found and downloaded it sometimes before. For me will be a very interesting function if the Zotero try to find related documents in my database automatically. It can be done manually by searching for title, doi, ISBN (or another unique item) but the automatization of this will be welcome.