Mostly I am interested in characters 32 through 126 (space, special characters, numbers, and upper and lower case letters) but the others would be good to know as well, especially the del, backspace, tab, and newline characters.
I have been told that there is a huge difference in the answer for different situations such as writing code versus writing a novel. Due to this problem I would like to use the average case for a large number of people with a wide variety of usage. Perhaps the same criteria that the people behind Dvorak used.
I plan on experimenting with chorded keyboards and I wanted this data so that I could assign the most common characters to the easiest chords and assign the least common characters to the most difficult chords. I don't actually need it yet since I haven't found how to intercept and replace keyboard events yet. In theory I don't need it at all and I could set all the chords at random but I am a bit of a perfectionist and wanted to optimize and generalize it. The reason for ASCII characters is that it should ensure complete coverage barring the use of additional character sets.