Are you suggesting is that FSA is giving unnecessary /wrong/poorly researched advice to the public?
Answering your last question first, pen: yes, I do think the FSA is giving unnecessary/poorly researched advice to the public.
Do you think there is space for any organisation/quango to filter/review scientific research so that the 'good & well informed'outcomes from research are properly mediated to the general public?
If yes: who should it be?
If No how will the general public know what research has found out?
This set of questions is harder for me to answer. I'll have a go. I think all medical trials should be published, even (possibly especially) ones with negative results as well as the ones with positive results because otherwise researchers and policy wonks (not an insult; I just can't be arsed to go looking for a different word; that one sprang to mind) cannot accurately judge what advice to give to the medical and related professions. Ben Goldacre, being far more knowledgeable on the subject, argues this far better than I can.
Based on that, then I see no harm in the publication of all scientific experiments, at least the ones paid for by the public. What I see harm in is rubbish conclusions being drawn from badly conducted experiments or those done with far too small a sample or those which, as in the roast spuds and burnt toast examples, are based on unnatural amounts of the 'harmful' substance being given to rodents and then the necessarily ridiculous results being extrapolated as applying to human beings who might occasionally ingest miniscule, harmless amounts of the same substance, and come to no harm whatsoever. The spikedonline podcast I refer to claims that this is what has happened.
They also argue that it is not the job of the FSA to nannify our lives over very small insignificant risks. I agree.
If you listen to the podcast you might also be convinced, or at least see where I and the podcast speakers are coming from, in thinking that quangos like the FSA have gone far beyond their original andnuseful remit and now seem to justify their existence in ridiculous ways. The podcasters argue that there should be funding reviews of such bodies annually, where they have to justify the importance of their next year's (or years') work to the health of the nation. That sounds to me like a good idea. I din't know who would do the reviews, but presumably there do exist bodies capable of such stuff. I'm thinking of NERC, for instance.
I hope that answers your questions. The podcast will probably answer them better.