ClinGen indels w/different reference bases

Clingen symbolic variant - If you have an indel or symbolic variant - and you send off the g.HGVS ie NC_000001.10:g.7682001_7685000del (which doesn't include the reference) to ClinGen and then you get it back - you will miss your variant as it will look for wrong ref

I guess the question is - do we count for allele/clingen purposes a different reference base as the same?

Or, should it in effect have been a failure (due to wrong reference base) - and we store it as an error?