Quickest way to deal with that is to download a book with a fair submission
from step 2 list and reject it with the kind of message I suggest from a
staff member. Something like "sentences repeatedly cut off in an orderly
manner. Usual cause is not holding down binding." Folks the reasons a
book is truly in need of rejection rather than a thorough going over with a
new version of k1000 can be easily made into informative messages.
I think a person submitting who has a version of k1000 should be asked to learn the optimize feature. It requires pressing one key and accepting the suggestions of the software. It radically improves scans. We need more ways for submitters to learn how to do things well, encouragement for those who do and rejection for some of the things we as validators are being asked to fix.
If someone gets repeated notices that their fair rated books are being rejected, we will not have to outlaw fair submissions. They will decrease. The stuff truly worth fixing will be fixed.