That's what the scraper scripts are for. For each site that does this, I have a bit of code which visits the article URL and pulls out the full content.
Is anyone aware of any repositories where the customization required to obtain the important content is maintained by a community?