Your interpretation of these data is misleading at best and dangerously misleading at worst.
>The risk of death from 1 hour of cycling reduces life expectancy by about 24 minutes. Wearing a helmet probably changes this by a couple of minutes.
This data does not take road conditions into account. Most cyclists prefer to bike on roads with very few cars on them. For example the road outside my apartment, a two-lane road outside of town by a lake with no services, is always packed with cyclists. The boulevard connecting my road to the train station, however, is heavily car-trafficked and has almost no cyclists. It is wrong to apply data from cyclists on the first road to those biking on the second.
>1 additional hour of sitting and watching TV in the evening, in an already sedentary life, probably has a similar effect on life expectancy.
http://well.blogs.nytimes.com/2012/10/17/get-up-get-out-dont....
This is double-counting. If you count sitting as a detriment, you can't turn around and count biking as an increment. There has to be a default from which to decrement or increment.
>[first link]>Every Minute Of Exercise Could Lengthen Your Life Seven Minutes
>[second link]>In fact, exercise was a bigger factor than body weight in many cases. People who were normal weight but were inactive actually lived an average of 3.1 fewer years than obese people who kept up high levels of activity.
This is an unjustified assumption of linear behavior. It's possible (likely IMO) that the first couple hours of exercise make a big difference coming from inactivity, but additional exercise on top of this likely has a smaller impact. The data only strongly indicate that cycling will have such a large impact on life expectancy if you already get very little exercise -- less than 2.5 hours/week of brisk walking. It is hard to believe, for example, that biking for 4 hours per day would increase life expectancy by 13 years -- there are practically no interventions known to have such a large effect!