apparently there are two different issues involved: on one hand, CHD-CDs created with very old chdman got metadata updated in this passage and this changes the resulting sha1.
this is sort of correct, in the sense that these old CDs miss pregap/postgap data and should be anyway marked as bad_dumps until they can be redumped. in other words, it is the old sha1 which was wrong in a sense (but the new one might be wrong too, if the disks need gaps data, so in both cases they should be considered as bad dumps)
otoh, the issue with laserdisks (and apparently gdroms too) is something different still under investigation
so yeah, there is something not very stable at the moment in the code, but I'm not sure about how bad the situation is