This comes up every now and then, and the moderators' response so far was to focus on the content. Because of country restrictions, IP restrictions, publisher's cookie policies, monthly free article quotas, per-article exemptions, corporate (or personal) subscriptions and availability of the article through Google search (or "web" link) it's hard to say what percentage of HN readership can see the contents.

The worst that can happen is that articles from a heavily paywalled site (ft.com, nature.com, theinformation.com) don't seem to get upvoted much, meanwhile articles from not so restrictive sources (wsj.com, nytimes.com) do make it to the front page occasionally, so I'm assuming a good portion of those votes come from readers actually capable of reading the full contents.