feat: simplify kademlia topology build up

[istae]

PR removes the overSaturationPeers check in connectNeighbours, and the second iterator neighborhood connector. We have more PRs up now that prunes oversaturated bins (which this PR causes to happen) and more optimizations to increase build up time.

This change is 

[mrekucci]

In general looks good but some code this PR deletes (e.g.: in the connectNeighbours) were quick & dirty optimization done before the 1.0 release which took topology buildup on our staging from ~10 min to ~2-3 min. Were these changes benchmarked for regression?

Reviewed 1 of 1 files at r1, all commit messages.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @acud and @metacertain)

---

pkg/topology/kademlia/kademlia.go, line 315 at r1 (raw file):

		if po < depth {
			return false, true, nil

Unacassarry new line.

[istae]

We benchmarked this on the main-net and there was no significant difference in build up time. We have a new branch now which optimizes kademlia on top of this simplification which should result in quicker build up compared to master.

Reviewable status: complete! all files reviewed, all discussions resolved (waiting on @acud and @metacertain)

[istae]

Some notes about the tests for this PR:
We ran all kademlia related branches, and master branch on the main-net, on bee-storage node 0, with preexisting state-store.

The /topology endpoint was queried on a 30 second interval for 20 times for each branch:

On simplify-kad branch: depth 12 was reached by min ~5
On prune-outofdepth branch: depth 12 was reached by min ~5
On balancer-opt branch: depth 11 was reached by min ~5

On master branch: depth 11 was reached by min ~4

On the master run for example, depth 12 was never reached, because of small population size in bin 12.

prune.zip
opt.zip
master.zip
simplify.zip