At risk of going round in circles, I had some concerns that thirdcrank's link is to a publication by Colin Clarke, which is a commentary on the Government-commissioned report by a known helmet sceptic (at least in my understanding). Even as a sceptic myself, I felt it better to read
the original report first. However, that worried me because of its apparent overriding of the scientific process. That is one of publishing research openly to the research community for comment and feedback.
The report just takes all relevant reports and assesses them, without evidently taking account of any later responses. That gap is somewhat filled by Colin Clarke's commentary, which does do that, although I'm left wondering about other responses to the same papers, which may have been more or less positive.
Of course, there's a quality test first in getting published at all, but that's only the first gate. The second and essential gate is that others look at your research, comment on the methodology, and try to do other tests that support or contradict your conclusions. So reading any published paper in isolation from subsequent responses is dangerous. Broadly, anything that gets no responses is not being seen as important, whilst anything heavily cited later is being accepted (or undermined, depending on what is said about it!) by the research community. Indeed, the level of citations is therefore itself used as an assessment criterion for research grants.
All of which is why of course the picture is confusing - there doesn't appear to be much consensus, and governments (and the media) are leaping on reports the moment they are published, without allowing the community review process to take place.