Jon Peltier discusses the purpose of charting, with a bad pie chart as the exhibit on trial.
For an excellent explanation of why pie charts are almost universally a poor choice of visualization see Stephen Few’s Save the Pies for Dessert (PDF).
I don’t quite agree with Jon’s contention that charts are often used to make presenters look smart, conceal the facts, accentuate the positive and so on. We use the tools at hand, and we use them as quickly as possible. When the tool at hand for data visualization is Excel, and the default settings are poor, and there’s no quick & easy access to advice, then the end product is bound to be unsatisfactory — and the user might never realize that.
The solution is better tools, made pervasively available. When every analyst and presenter can reach out and grasp a superior tool for the job of information visualization — no learning curve, excellent default presentation, useful advice when needed — then we will see better analysis and better decisions.
Open Analysis 
Heh heh, I was being a little sarcastic. (WordPress doesn’t have a filter for that, only for messing up my code samples.) Most of the distortions in my list are not pursued maliciously but through lack of knowledge, but as Heinlein said, “don’t rule out malice.”
So how do we reduce these distortions? Better tools would be a large step in the right direction. I don’t believe you’ll ever have a visualization tool with “no learning curve”. The tool itself might be easy to use, but there still will be issues getting the right data into the right form, choosing the right style of chart, etc.
Without learning about the effectiveness of different chart styles, and without resisting graphical capabilities that include distracting and distorting colors, shadows, and other gratuitous visual effects, we will continue to see ill-designed charts that have led to criticism like Raymond’s, on which I based my post.