Make info logs easy to reason about

[frol]

I suggest we print a regular info log as a stats about all the blocks that are currently getting indexed (similarly to how nearcore does it), where we include:

• List of the block heights that are processed (if there are too many of them [let's say 5], print 4 oldest ones and 1 the most recent)
• Indexing speed (blocks per second rate)
• Estimated time to reach the tip of the network (the most recent synced block height - the most recent indexed block height / (indexed blocks per second - synced block per second))

As a note, we can use BTreeSet to maintain the list of currently processed block heights.