The article talks about a million web pages, but it's actually the top million domains. This is an important difference because the top sites could be completel

The article talks about a million web pages, but it's actually the top million domains. This is an important difference because the top sites could be completely different to the web pages that we view. For example, Facebook is probably high up the list. But which Facebook page do they measure? The 'front page' of a non logged in user, presumably? That would be vastly different from the majority of FB users. Likewise, Wikipedia is also an extremely popular site. But surely most users are looking at specific pages of interest, and not just the front page?