I’ve been involved in dozens of these investigations over the years, and it really is tough. Data is generally sampled - not continuous, and tracing the problem through the data stream to the initial source is very tedious. We actually used the color of the plumb from photos as the final key in one investigation.
And the pressure is great!
Good luck to the team.
I'm guessing you meant "color of the plume."
Didn't SSMEs keep running for something like 500 mS after the Challenger explosion?