Add clustering to BigQuery tables where appropriate

[medvedev1088]

https://cloud.google.com/bigquery/docs/clustered-tables. It can make some queries cheaper and faster.

[allenday]

I was thinking of this when reading @askeluv 's blog post: https://towardsdatascience.com/how-to-get-any-ethereum-smart-contract-into-bigquery-in-8-mins-bab5db1fdeee If we implement clustering/partitioning, it should be possible to reduce the query cost of using the logs table directly, i.e. no need to create contract-specific tables.

However, there is some active work being done for streaming into partitions, see: https://issuetracker.google.com/issues/35905817#comment89

[medvedev1088]

That's a good idea.