Until seeing Tufte's criticism, I was quite satisfied with the options presented to me. The weather tells me the one thing I'd like to know, sunny, rainy, cloudy, and so forth, along with the temperature. I don't really want to know the humidity. The basic information of sunny/cloudy is enough for me to make some decisions of what I'm going to wear. Weather predictions are bad enough anyway. There have been times when it hasn't done me any good, where as the days pass, the weather predictions change for the following days. So unless I get a degree in meteorology, I don't won't be compiling barometric pressure, humidity and precipitation measurements to decide what to wear.
Further argument against the need for more information, is the stock market page. I will not need to know the accuracy of the numbers to a 2 significant digits for every day of the past 3 years. The general graph of up and down is well enough. Again, that really won't help with predictions though, because stocks are not as predictable as they are sometimes presented to be [read Nassim Nicholas Taleb].
I do agree with Tufte's closing statements from his critique, "to clarify, add detail... and clutter and overload are not attributes of information, they are failure of design." The alternatives offered so far, just do not compel me as being clearer graphical solutions combined with information valuable to me.