Is there a significant difference in value of a backlink from a unique vs. non-unique article/website? Common sense tells me that there should be, but has anyone tested this somehow? And would spinning the text help to make it unique?
I don't believe it makes a significant difference at this time, but it could in the future. I believe that it does make a difference some times but it's a random algo feature that is dialed in and out.
That's a different issue altogether, imo. The efficacy of that may be affected by the sites that publish the reproduced content. Crap inbound links subjected to depreciation, funky link graph, etc., etc. That's an issue with the sites that are publishing the content, not the technique itself.
I will give you an example of a technique working because of positive signals. I believe this loophole or bug in Google's algo is one of the reasons why copied/stolen content on decent sites can outrank the original documents. Google is making decisions based on quality signals but that allows the technique to work as long as the web pages have enough of the positive signals and lack the negative ones. I just removed a long post by someone that was posting the stolen content across a lot of forums. His documents and those of others who have republished the content are outranking the original publishers. That's a loophole and a bug in the way the algo works, imo.
Till date I have not seen any negative impact on my websites by submitting the same article on more than one site. Its definitely a matter for the websites publishing the duplicate content, but there is no harm for the site getting the back link.
But a common sense says that if the publishing site is not getting the real benefit of the duplicate content, the site getting the back link from that page is not getting the actual benefit.