[stacks-signer] Update signers to query stackerdb for DKG messages to prevent DKG stalls due to timing issues

[jferrant]

Suppose we have three signers, signer-0, signer-1, and signer-2.

Signer-0 and signer-1 receive burn block X which happens to be in the prepare phase of the next reward cycle, but it happens to be right in the middle of their nodes calculating the registration set so they do not see themselves as registered.
Signer-2 gets burn block X, but the milliseconds later is all that is needed for the node to finish the registration calculations. so it finds itself as registered AND that it is the coordinator. So it triggers DKGPublicGather sendings its public shares to other signers.
Signer-0 and Signer-1 receive this public share but because they ONLY check for a refresh once a burn block arrives, it ignores this event until burn block event X+1 shows up and THEN it queries for the registered signers. However, it never responds to the initial coordiantor message as it was ignored when it first arrived since they did not think themselves to be reigstered.

Fix this.

This may also fix #4630
This will also address #3990

[blockstack-devops]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.