I've asked it too, no replies. the common reason that people bring for things being like that is that a sequencer usually behaves stopping his clock on stop.
WELL
this is awful for MIDI
analog clock stopping = no problems at all. it's analog, and as soon as there's signal, the cv's are triggered.
but MIDI clock= disaster. if the slave misses the note on signal before the clock start, it simply doesn't play. if the clock is sent after the note, the tempo is shifted, creating jitter.
WHAT DOES IT TAKE? in MIDI world, the start-stop commands can be sent as Machine control (realtime sysex) AND cc controls. nothing exoteric. PLUS - it's there already.