Created byte_estimate_to_str(), much like byte_count_to_str(), except the
byte count is rounded to the nearest tera/giga/mega/kilo-byte
(technically tebi/gibi/mebi/kibi-byte, but ugh, I'm old-sk00l!).