If you want to know more about the quality and accuracy of the ECMWF ensemble forecasts of wind speed, you should have a look at a paper that my collaborator Renate Hagedorn and I just got accepted for publication in the journal Meteorological Applications. We look there at verification against analyses and observations at app. 700 synoptic stations over Europe… The good news is that these forecasts are really good! The maybe not-so-good one is that verification against analysis or observations gives very different pictures of the forecast quality.
The preprint of that paper is available on this site… I will post a few figures from the paper soon!