chore: minor improvements + fixes

[JakubFornadel]

Purpose

This PR fixes 2 things:

• we will now catch up with the rest of network and get within 1 step. Node, which is catching up with the network is increasing its steps 40x faster then the other nodes. This means that if network is stalled for lets say 24 hours, new node will catch up with the networks in about 35 minutes. This is defined by kMinLambda and kMaxLambda. Right now kMaxLambda(60s) / kMinLambda(1.5s) = 40x. We can increase this by increasing kMaxLambda
• we were not saving and proposing own votes before, which was a pretty important bug -> it is fixed now. Lets say if there is network of 4 nodes and 2 of them go offline. The other 2 proceed to step 100. The other 2 go back online but because we never broadcast current round next votes, the other 2 have to catch with the network (this finally works) but if we were broadcasting also our own next votes for current round (in case we are stalled in the same round), we would be able to move to the next round much sooner.

How does the solution address the problem

Changes made

[taraxadeploy]

prnet cleaned-up successfully