[Feature Request] Heavy remote codecs need longer timeouts

[cretz]

Describe the solution you'd like

There is a use case where a slow remote codec can timeout on workflow show. The Go SDK defaults to 10s for get history events call for each page, and that 10s is for the context for the whole gRPC call and remote codec is implemented as an interceptor within that.

The beta version of the CLI and tctl had a --context-timeout that allowed users to increase the timeout (which the Go SDK respects because by using max(1s, user-timeout / 2)). We can add this in back in (though would suggest --rpc-timeout) or find other solutions.

We can also consider an option to concurrently call the codec with all payloads needed instead of serially in a visitor like we do today, though that's not necessary for this issue.