The same way that you can look at the graphs in your examples and believe they in any way mean helmets have little protective capabilities. I will take just the first 2, as they are written up similarily.

The graphs in each of those say simply one thing. That Bicycling is marginally more dangerous than walking. Beyond that, anything else is pure conjecture. How you get from those trends to "By ALL means feel skeptical and ride as if a helmet would offer you no protection at all since it is unlikely that it will." is a stretch.

First off, those graphs say nothing about the causes of death. I will make the assumption that they are primarily, if not all automobile impact related. Only because I will assume the primary cause of pedestrian death is also impact with an automobile. I would agree that getting hit by a car will usually cause a fatality, helmet or not. That is simply laws of physics. In a crash with a car, a cyclists is basically the same as a pedestrian, as the bike affords little to no protection.

A useful graph would be one to break down the cause of cycling fatalities and their trends.

Here is an example of an accident where a helmet saved a life, namely my brother in law. And I would consider this more than a "minor" accident, the kinds you state "the protection it does offer is that in minor accidents you will save yourself a minor injury"

My brother in law hit a rock while riding, I don't recall the details on speed or location. He went over the handlebars and landed on his shoulder/head. Broke his collarbone. Upon getting up there was a sizable, sharp rock embedded in his helmet. Had it not been there, that sizable, sharp rock would have been embedded in his head.

I will keep my helmet on...

-D