Reconsider single commit propagation

[shuse2]

Description

Re-consider how single commits are handled in the pool, and how they are handled in the P2P event handlers so that the usage of event handler and validator registered here will be more standard usage

Motivation

Currently, single commit propagation is handle in validator in the p2p not to re-publish immediately after receiving the single commits, and on the actual handler, it does not do anything.
In general, this is not intended way to use the validator, so we should avoid this pattern.

Additional Information

Related PR #72

[shuse2]

There are 103 single commits and at the moment we send single commits to 16 peers. If we directly forward one-by-one, nodes will receive single commits 16 * 103 = 1648 times in expectation per BLOCK_TIME. That is why you wait to receive more of them and then bundle them together.