I had not suggested summarizer should make any special effort to punch through paywalls.
Rather, it should be obvious content like this is not the actual document that summarization is requested for. This could be useful even in the case someone has a subscription to a site and just sent the URL across, forgetting or not realizing that Summarizer would not see the same contents. Hence "if summarizer hits a paywall it should probably just say it can't read the article because it encounter a paywall instead of summarizing the paywall page."
