chore: rpc usage metrics

[mfw78]

Background

It's difficult to rationalise issues associated with RPC connections and the watch-tower. More metrics are needed around this.

Details

Add additional metrics that are able to comprehensively track all RPC usage throughout all code paths.

Acceptance criteria

The following metrics should be implemented:

• RPC calls (eth_call)
• Gas consumption for (eth_call)
• Time elapsed between new block notifications
• Time elapsed processing a block (ie. from receipt to polling finished)