I tend to use what I call a bootstrapping method for doing literature reviews. I start with a paper that is relevant to the topic area. I then work backwards and forwards from that paper. I pull all the relevant papers that the paper cites and use a website (doesn't really matter which one) to find all the relevant papers that cite the paper. I then repeat the process for each of those papers. As my literature search is nearing the end, I try and look at the webpages of the key authors in the area to see if they published something obscure.
The reason the website does not really matter is that by going backwards and forwards you are leveraging the literature searches that were done for all the other papers. As you are not searching for key words, but rather by citations, the search engine is not as critical. What matters is if the relevant papers are published in places that the search engine indexes. If something is not indexed in the major academic databases, is not referenced by other authors, and was written by someone peripheral to the field, then, unfortunately, the work will probably get missed.