Can't help on the first one, but the Bouquets behaviour you've described is the same in C10d, so must be inherent in the current CVS Neutrino design.
Enigma keeps to the current bucket, though, so maybe neutrino could be tweaked to do the same?
Cheers- AL