[ethcore] Untangle author, engine signer and reporter

[dvdplm]

The code in ethcore has grown organically over time and the --engine_signer and --author CLI arguments end up in the same struct field from where it is then used for different things depending on the engine used. The terminology used is also a big mess with "signer", "author", "sealer", "engine_signer", "consensus_signer" or simply "address" being used to refer to the same thing.

In the Aura engine, the --engine_signer address is used for two things:

• sign blocks when the node is part of the validator set
• sign misbehaviour reports

As pointed out in this comment it might make sense to use two different fields, e.g. consensus_signer for block production and report_signer to sign misbehaviour reports to make it clearer in the code when we perform one or the other function.
All nodes provided with a report_signer address can submit reports, but only nodes with a consensus_signer can sign blocks (if/when they are part of the validator set as determined by the engine).

So this ticket has two goals:

1. clean up naming and disentangle the CLI arguments;
2. make it crystal clear in the Aura code which address is used for which purpose: block signing and misbehaviour reporting

In theory the work could be split into two chunks, but in practice I think it's easier to address them at the same time.

[adria0]

Closing issue due to its stale state.