fix: avoid wasteful reprovides outside threshold

[2color]

What

This PR fixes a couple of DHT bugs related to how provider records are handled and processed:

• fix: bug in the logic in packages/kad-dht/src/reprovider.ts that causes wasteful reproviding of records that long before reaching the threshold. This would cause valid records to be provided every REPROVIDE_INTERVAL
• fix: don't delete expired records if they're ours to ensure we persist the user intent to provide CIDs even if the node goes down beyond expiration. These will be reprovided in the next processRecords run
• fix: return records within PROVIDERS_VALIDITY and delete MAX_RECORD_AGE which seems to be from before the increase of record validity to 48 hours, and is in essence an outdated duplicate. This fix ensures that provider records are returned in RPC requests as long as they are within the 48 hour PROVIDERS_VALIDITY.

Other small changes:

• Rename cleanUp (which is misleading given that it triggers the reprovides) to a more descriptive name processRecords
• Add a few more logs and reduce the level of logs from trace to normal logs

Notes & open questions

• For self provider records, should we check the block store to ensure they haven't been garbage collected or removed?

Change checklist

• I have performed a self-review of my own code
• I have made corresponding changes to the documentation if necessary (this includes comments as well)
• I have added tests that prove my fix is effective or that my feature works