Misleading log message for lost runs

[taylordowns2000]

The log message displayed for a run that has been lost is misleading. Rather than showing "Nothing yet" we should have a separate log, input, and output message for when a run is in one of the :final_states.

Desired wording:

1. Logs: "No logs were received for this run."
2. Input: If no other data_retention related messages are shown, then: "No input state could be saved for this run."
3. Output: If no other data_retention related messages are shown, then: "No output state could be saved for this run."

Note that we've now got a lot of conditionals around what to display for dataclips. How do we know that an output dataclip was not saved because of a crash, instead of not being saved because of a customers input/output data storage policy?

[taylordowns2000]

@elias-ba , we are trying to re-articulate this together: 80/20...

• For logs there is a fallback which is currently "nothing yet". We want to make that fallback conditional on the state: if the state is in final_states make the fallback say "No logs were received for this run." if not, make the fallback say "nothing yet..."
• For input and output, there are currently 3 states (real JSON data, dataclip wiped, nothing yet...) and similar to logs, we want to adjust the fallback state from always being "nothing yet" to being "No ${input/output} state could be saved for this run." if the run state is in final_states.