Quite a few helmets around that are designed to reduce rotational forces, of course the standard certifications do not bother with such measurements and there are only a couple of independent testers covering a small sample of the market who actually publish detailed numbers.
As far as the recording guidance goes, any broken bone (or suspected break) is supposed to be recorded as serious though it's generally considered that a number of serious are misreported as slight and a large number of slights aren't recorded at all (potentially up to twice the number that are)There are only ~4000 cyclists seriously injured each year, and not all of those will be head injuries (I'd say a broken leg is serious for example).
On the flip side, as far as judging transport risks goes, the collision stats will include and be biased by sportsing road cyclists who, at least in my anecdotal impressions, appear to be the group a great many of the 'my helmet saved my life' advocates belong to and generally appear to crash more. That's a very different risk profile to someone cycling to the shops.
This also shows if you use the cycling distance estimates from the LTDS with the casualty figures from the greater London area, the fatality rates are significantly lower than the national average, despite the ability of London news of cyclists being crushed by HGVs being better able to make national significance.