It seems that the consensus is to just fix StringUtils.replaceChars(S,S,S), so I have done so.
I have added a couple more unit tests to cover more of the Javadoc examples and the example cited in the email thread. Please note the following: the pre-fix unit tests asserted that if an input string was not "modified" (in quotes since Java String are immutable), the return String had to be the "same" (Assert.assertSame as opposed to Assert.assertEquals). The current impl return a new string from the working buffer even if no replace took place. The Javadoc contract does not specify the old behavior or even hint at it, the only mention of returning the input string is if it or the search chars are empty. Please let me know if you think this is a problem. Thank you, Gary
