The method returns a float because the max. number of bytes/character could actually be fractional depending on the method of encoding.

If future encodings had fractional bytes/character instead of say, UTF-8's 1-4 bytes/character, you may have been thankful.

So to answer your question, yes I would assume. As long as you know the encoding scheme and whether or not fractional bytes are being toyed with, you can make the decision yourself.