CHAIN_HEAD_NOT_AVAILABLE ought to be a warning, not a trace log

[matthew1001]

It's not difficult to cause CHAIN_HEAD_NOT_AVAILABLE to occur in a heavily loaded validator. I'm not sure exactly what causes the node not to have a chain head available at a given point in time, but when it happens it causes a nonce gap for the sender of the transaction.

Since the log entry rejecting transaction {} due to chain head not available yet is a trace entry, you can only spot that this is the cause of a nonce gap by recreating the issue with trace enabled.

I think there is an argument for turning it into a warning log entry. If it happens so frequently that the warnings make the logs noisy, there's probably something the user needs to look into anyway.