On KGW, that should take you about 15 minutes if you are really interested.
To start I said I wasn't interested in KGW. But nevertheless, that was interesting and after reviewing it I think I understand where the exploit in KGW lies (and no it isn't what you suggested, exactly, nor is it merely "adjusting too fast"). But in any case the same exploit doesn't exist in cryptonote coins. Of course that doesn't rule out a different one.
If I have time I will write up what I think I found in KGW more carefully and see if it really holds up. Perhaps some of the coins using it will find that interesting.