Since 2012, the Semantic Web journal has been accepting papers in a novel Linked Dataset description track. Here we motivate the track and provide some analysis of the papers accepted thus far. We look at the ratio of accepted papers in this time-frame that fall under this track, the relative impact of these papers in terms of citations, and we perform a technical analysis of the datasets they describe to see what sorts of resources they provide and to see if the datasets have remained available since publication. Based on a variety of such analyses, we present some lessons learnt and discuss some potential changes we could apply to the track in order to improve the overall quality of papers accepted.