Well, something to share with you is that I just finished my listening tests making on purpose a 2mm overhang error and other test changing only ( what more or less permits the headshell ) the angular offset. I think you need to do it.
If you only change overhang without a corresponding change in Zenith you will not be in a good place. Here is a standard Lofgren A compared to a Lofgren A with a 2mm increase of overhang.
here is the same +2mm with a complimentary 1.4° change in zenith.
and to complete things here is back to the standard Lof A overhang with a 1.4° Zenith error.
dave