I'd be interested to know if there are any cases where RHEL have issued a WONTFIX on a bug due to some software relying on the bugged behaviour. Given how much of it is free software, would they really not merge a fix into their releases to keep the bugged behaviour?