> Check Point wanted to tamper with the code responsible for calculating hash block signatures or for comparing segments of hashes with verified ones to trigger an exploit.

So fundamentally it's a parsing exploit on untrusted user input. They should have been able to simplify and test this mechanism extensively enough to not have this type of vulnerability, because it doesn't sound like researchers even had to go to the level of the timing attacks that came up recently in a related area [0] [1].

[0] http://tpm.fail/ [1] https://news.ycombinator.com/item?id=21520074