Couldn't you (gently) use the wayback machine for this? Or just start crawling feeds with something like scrapy? There isn't enough RSS, I feel, to make this prohibitive.