This study was subject to revisions - given the author inferred conclusions, as opposed to emphasising estimates only.
http://www3.imperial.ac.uk/newsandeventspggrp/imperialcollege/newssummary/news_17-7-2014-8-56-10
For example (I have underlined my thoughts);
"
Comparing the only data available, the team estimate that ten times more fires are happening than are being reported. Instead of an
average of 11.7 fires each year, which is what is reported publicly, the researchers estimate that more than 117 separate fires are breaking out in turbines annually."
Consequently, a focus should be on improving data collection and retention, as opposed to making "estimates" which are significantly above actual reported data. These estimates distort the actual (based on the current data) and allow for sexed-up comments like seen from the original poster.