prior discussion on COSMOON slash: https://kusama.polkassembly.io/post/490
it's my opinion that ParityDb should already be understood to be experimental, but it's prevalence in communications and use in public benchmark statistics leads me to believe that there are likely significant numbers of independent validators using it in production already.
Since this slashing event has caused a revision in CLI flag names and explicit recommendations, it is probably reasonable to continue as if prior to this event the risk was not well-enough understood.
I am putting forward a motion to cancel the slash of COSMOON based on that assumption.