Changing CentOS in mid-stream

There's some truth to that, but fortunately my experience has been very positive. First, I haven't find it that hard to minimize the time spent on upstream reports that aren't interesting. That probably varies a lot depending on the project--more popular and user-exposed projects probably require more triage.

Second, some upstream bug reports have been very useful. Either they've caught things early, or hit on a reproducer for a serious but rare crash, or just invested a lot of work in investigating the problem. In the upstream NFS world we've been lucky to have some really outstanding bug reporters.