duplicated revision number information in the consumer-addition proposal

[AaronKutch]

When trying to launch a consumer chain on testnet, we made a last minute change to the chain id which added on -testnet-1, which I didn't believe would possibly break anything. However, I discovered that Hermes could not connect the chains for some reason, and after a lot of hair tearing out and chasing why certain chain ids were causing obscure trusted header mismatch errors, I found that the "revision_number": 0 subfield in the consumer proposal clashed with the revision number associated with the "chain_id" . Different things were using different revision numbers and making the tendermint clients fundamentally incompatible.

Either this field should be deduplicated, or the proposal command should have some kind of check to prevent revision number disagreements.