Last week I was on a committee to give out a “best paper” award. Of the six nominated papers, two contained graphs that were completely incomprehensible because they had been produced in high-res color but were given to the committee as photocopies. In fact, I think this was 2 out of 2 of all the papers in the competition that had graphs. I’ve seen this happen a lot. I myself have an AJS article whose gray scale graphics, while legible in the original printed journal, are completely illegible in the low-res JSTOR version, or any photocopy you might make of that article. Having been bitten badly by this problem, I am very attuned to it. Printing on a low-res printer has the same problem as photocopying. Lots of journal reviewers print out the papers they are reviewing. I personally use “fast draft” to save ink and time when I’m printing most stuff. What do you use?
Advice to the wary: Microsoft standard colors and gray scale are just not going to be distinguishable after photocopying or being printed on a low-res monocolor printer. Some of the colors will be indistinguishable from white, and even the difference between black and light gray will disappear in a photocopy. Simple line graphics with markers are much better.
If you are sure you only want your work read on line, you don’t have to worry about this. Otherwise, the word to the wise is test for readability under adverse conditions.