I'm afraid what we have here is the usual mismatch between 'common sense' and 'actual events'.
There hasn't been an introduction of any compulsory road safety devices that have led to a decrease in serious injuries in the following year (as time goes on, it's harder to compare like-with-like as other safety factors come into play). Crash helmet legislation for motorcyclists was followed by an increase in both deaths on motorcycles and head injuries, seat belts compulsion was followed by a rise in serious injuries and deaths in car accidents, and I'd be willing to place good money that if cycling helmets were made compulsory then a similar pattern would emerge.
Although poor design is a large part of the problem - crash helmets in particular have lead to para and quadraplegic injury increases since compulsion due to the likelihood of the head snapping back and taking the spine with it, an issue that wasn't comon before legislation - the bigger issue is that of what is known as 'risk compensation'. This theory suggests that everyone has a built in level of risk they are willing to take, and the safer they feel, the greater the real risks they take even if the 'safety devices' used are little more than a placebo.
Look at it this way: you have a big sports car. No one tells you to wear a seatbelt, it's entirely your choice. You need to get up a lot of country roads to get to your destination.
Are you likely to drive the same way with the seatbelt as you would without it, or are you likely to be more careful without? If you say you'd drive the same, then either you're in a tiny minority, or you're lying.
Exactly the same argument applies to motorcycle helmets. I've done a lot of miles without a lid on, but a huge amount more with. I know which scenario I'm most likely to take risks in, even though I know that a helmet isn't going to save my brain from getting turned to mush in a head on collision, any more than my skull would. I ride flat out (a massive 65mph....) with my lid on, but rarely have I reached 45 without it.
I don't see cyclists being any different.
BTW The injury comparisons at the stgart of this post are per x miles, not a sum total.