Round tripping through conversions to other codesets was always only temporarily needed, and for most non-Unicode codesets it was always limited to one or a han

Round tripping through conversions to other codesets was always only temporarily needed, and for most non-Unicode codesets it was always limited to one or a handful of scripts (e.g., you could not round-trip text that included CJK and Cyrillic through any non-Unicode codeset). So I'm not sure that having that feature was all that important. However, early on it was indeed helpful to obtaining adoption to be able to convert with simple lookup tables.