how to infer cutting sites from the result?

[Huanle]

Hi Crispritz developers,

Thanks for developing and bringing crisprtiz to the community.
I have a layman's question regarding inferring cutting sites from the predicted (off-) targets.
Basically, when a gRNA match to the same target region but with bulges introduced to different positions in the target DNA, will the cutting sites vary? aka, will the position of the bulge alter the 3nt-proximal-PAM-cutting-site rule.

The attached image contains one such example.
Thanks a lot in advance for your help!