fix: Use threshold number for previous running state in resharing

[ghost]

We have a checkpoint, wait_for_ready, in the running state that waits on a threshold number of participants are online before starting the running state. This check has two bugs:

1. This check is done before spawning the resharing handle, which is not the correct behavior. The resharing handle should be spawned independently of the starting conditions of the running state.
2. It should not use the threshold size of the new resharing state in case of resharings. The node needs to wait on a threshold number of the participants from the previous node set.

Both issues are fixed in this PR. This is consistent with the behavior previous to merging running and resharing state in the node

mpc/node/src/coordinator.rs

Line 378 in 83e0970

sender

[kevindeforth]

Good catch!