Outbounds with a round 7 failure should be rescheduled to the same vault instead of dropped
There was a round 7 failure reported in https://thornode-v1.ninerealms.com/cosmos/tx/v1beta1/txs/ED393D7776D73F0545E4A08690FFE9F177699C3B705224C1A061BD8EC1E5CD2F. Since whether the failure was round 7 may be known by only a subset of parties, we err on the side of freezing aggressively: https://gitlab.com/thorchain/thornode/-/blob/v1.127.2/x/thorchain/handler_tss_keysign.go?ref_type=tags#L183-210.
This can result in a situation where > 33% of members mark the item in signer storage as a round 7 failure to retry indefinitely, but nodes that do not see the round 7 failure will only retry so long as it remains in the outbound queue. If the reschedule period passes and the vault remains frozen, this results in never being able to achieve quorum since some nodes remain stuck on the retry signing while others have moved on.