Massive unrealistic DUP in Sniffles 2.7 output

[Dario-Galanti]

Thanks for this amazing tool! Wanted to point out that I just switched from Sniffles v2.6.2 to v2.7 and suddenly had some massive duplications that are biologically unrealistic.
Some are as large as chromosomes, here is an example of a 29MB DUP (ripped of the alleles for simplicity), but I have even longer ones.
Scf9YQZ_11_HRSCAF_22 23271923 Sniffles2.INS.dup2F07SA . . 58 PASS PRECISE;SVTYPE=INS;SVLEN=29246330;END=23271924;SUPPORT=6;COVERAGE=15,21,20,17,11;STRAND=+-;STDEV_LEN=0;STDEV_POS=0;VAF=0.500;AC=1 GT:DR:DV:AD:DP:PL:GQ 0/1:17:6:17,6:23:.:12

I saw this is part of a commit of the new version:
372bc60
so I imagine it is intended behavior to some extent, but it is probably good to be aware of it.
It completely crushed my pangenome construction pipeline (vg construct).