the standard quote isn't directly relevant to this situation, but it's close.
I don't think the requirements to use the research paper publishers are due to malice, I think it's inertia.
In the days before the Internet, publishing was a very expensive thing to do, In such an environment an organization dedicated to separating the wheat from the chaff and publishing the wheat was an incredibly valuable service to provide.
Over the years, as the organization providing this service was able to make a profit from publishing things, and the cost of publishing has dropped, I think they have become less critical about what they publish, and so their value as a filter for 'the good stuff' has been dropping.
With the cost of publishing now almost zero, there would still be value in the service of evaluating papers to find the good ones, but I don't think any of the research publishers are really providing that service effectively anymore.
As such, I think that publishing the papers in a place where Google can find them (and apply the pagerank type algorithm to them) is at least as effective an indication of the probably quality of the papers.
It would be good if the various industry organizations would recognize this and make all the papers available, and provide a service for their members by reading everything they can and provide feedback to the author and quality scores for their members (along with indexing services to help their members find things)
I think that simply the process having a lot of people reading disparate documents would be valuable as the readers would be able to spot things across the different documents that the authors of the documents themselves are unaware of.