Could scientists ‘peer-review’ the web?